CH-13 A Second Look at Field Research – Research Methods: A Process of Inquiry : Exam Guide

Chapter 13 A Second Look at Field Research 13.1 Conducting Field Research 1) Naturalistic observation and case-study research allow us to A) identify contingencies B) prove causality C) manipulate the independent variable D) rule out alternative hypotheses. Answer: A Rationale: Naturalistic observation and case-study research allow researchers to identify contingencies, patterns, and relationships within real-life contexts. They provide valuable insights into behaviors and phenomena as they naturally occur. 2) Which of the following does NOT apply to conducting experimental research in natural settings? A) We often cannot obtain enough participants for the creation of experimental and control groups. B) We cannot test the external validity of causal conclusions arrived at in the laboratory. C) Random assignment is often impossible. D) Manipulation of independent variables is usually difficult, if not impossible. Answer: B Rationale: Conducting experimental research in natural settings does not preclude testing the external validity of causal conclusions arrived at in the laboratory. Natural settings offer opportunities to validate findings in real-world contexts, enhancing their external validity. 3) Which of the following is NOT a type of field research? A) field experiments B) program evaluation C) survey D) stimulus-response studies Answer: D Rationale: Field research encompasses various methodologies such as field experiments, program evaluation, and surveys. Stimulus-response studies typically refer to laboratory-based research rather than field research. 4) Which statement is most accurate? A) Research in natural settings is limited to low-constraint methods. B) Experimental research cannot be carried out in natural settings. C) Field research can be carried out at all levels of constraint. D) Field research is limited to behavioral observations. Answer: C Rationale: Field research can be carried out at all levels of constraint, ranging from low-constraint methods such as naturalistic observation to highly controlled experimental designs conducted in natural settings. 5) Why is naturalistic research NOT useful in answering causal questions? A) The participants are not representative of a population. B) There is no observational variable. C) Naturalistic research does not seek relationships among variables. D) It does not rule out alternative hypotheses. Answer: D Rationale: Naturalistic research is not useful in answering causal questions because it does not involve manipulation of independent variables and does not rule out alternative hypotheses, making it challenging to establish causal relationships. 6) Which of the following is most likely to be studied in a naturalistic setting A) the evaluation of the effectiveness of a new tutoring program at the University of Nebraska B) the effect of Thorazine (an anti-psychotic medication) on non-psychotic individuals C) the effects of alcohol on the mathematical ability of college freshman D) the type of neurotransmitter released during stress Answer: A Rationale: The evaluation of the effectiveness of a new tutoring program at the University of Nebraska is most likely to be studied in a naturalistic setting because it involves observing behaviors and outcomes in a real-world educational environment. 7) Which of the following is an example of field research? A) An evaluation of the effectiveness of public health programs. B) An evaluation of the effects of natural disasters. C) An evaluation of employee reaction to new work conditions. D) All of the above Answer: D Rationale: All the options listed involve conducting research in real-world settings outside of a laboratory environment, which is characteristic of field research. 8) What is the main reason naturalistic observation cannot easily answer questions about causality? A) Reliable measures cannot be obtained in natural settings. B) Valid measures cannot be obtained in natural settings. C) Independent variables cannot be manipulated and alternative explanations are not ruled out. D) None of the above; naturalistic observations easily answer questions about causality. Answer: C Rationale: The main reason naturalistic observation cannot easily answer questions about causality is that independent variables cannot be manipulated, and alternative explanations are not ruled out, making it difficult to establish causal relationships. 9) What is (are) the reason(s) that conducting experimental research in naturalistic settings is difficult? A) It is limited by demands and characteristics of the setting. B) Random assignment of participants to groups is very difficult. C) It is often difficult to obtain the number of participants required to create experimental and control groups. D) All of the above Answer: D Rationale: Conducting experimental research in naturalistic settings is difficult due to various reasons, including limitations imposed by the setting, challenges in random assignment, and difficulties in obtaining sufficient participants for experimental and control groups. 10) What is an example of an independent variable that is impossible or unethical to manipulate? A) Exposure to a natural disaster. B) Participation in group or individual therapy C) Living near a previously unknown chemical dump D) Both A and C Answer: D Rationale: Both exposure to a natural disaster and living near a previously unknown chemical dump are examples of independent variables that may be impossible or unethical to manipulate in an experimental setting due to ethical considerations or lack of control over the variables. 11) It is generally easier to obtain ________ validity in experimental research in laboratories and ________ validity in experimental research in naturalistic settings. A) internal; external B) external; internal C) construct; content D) content; construct Answer: A Rationale: Option A correctly identifies that it is generally easier to achieve internal validity in experimental research conducted in laboratories due to the greater control over extraneous variables. External validity is often easier to attain in experimental research conducted in naturalistic settings because the conditions more closely resemble real-world situations. 12) Which of the following is NOT a major reason for conducting experiments in field settings as opposed to laboratory settings? A) to test the external validity of causal conclusions arrived at in the lab B) to determine the effects of events that occur in the field C) to have greater control over confounding variables D) to improve generalizability across settings Answer: C Rationale: Option C is incorrect because experiments in field settings typically involve less control over confounding variables compared to laboratory settings. The other options correctly identify major reasons for conducting experiments in field settings. 13) Why do we conduct research in the field? A) to test instructional procedures developed in the lab in actual field settings B) to meet growing demands to test the effectiveness of a large variety of social programs C) to take advantage of the potential for greater generalizability of results to other conditions D) All of the above Answer: D Rationale: Option D accurately reflects the reasons for conducting research in the field, including testing instructional procedures, evaluating social programs, and achieving greater generalizability of results. 14) Experimental research conducted in the laboratory allows us to test hypotheses under conditions designed A) to increase the likelihood of publication. B) to maximize external validity. C) to maximize internal validity. D) to reduce scientific rigor. Answer: C Rationale: Option C correctly identifies that experimental research conducted in the laboratory is designed to maximize internal validity through strict control over extraneous variables. 15) A high degree of experimental control may increase internal validity but often reduces A) scientific rigor. B) construct validity. C) believability. D) external validity. Answer: D Rationale: Option D accurately reflects that while a high degree of experimental control can enhance internal validity, it often comes at the cost of reduced external validity due to the artificiality of laboratory settings. 16) The major effect of reduced external validity is A) difficulty in generalizing laboratory experimentation to the natural environment. B) a decrease in internal validity as well. C) a decrease in all types of validity. D) that it invalidates the study. Answer: A Rationale: Option A correctly identifies the major effect of reduced external validity, which is the challenge of generalizing findings from laboratory experiments to real-world settings. 17) Testing the external validity of causal inferences that have been found in high-constraint experimentation A) is one reason for doing field research. B) must be carried out in high-constraint laboratories. C) cannot be done. D) must be restricted to correlational research. Answer: A Rationale: Option A accurately identifies that testing the external validity of causal inferences derived from high-constraint laboratory experiments is one of the reasons for conducting field research. 18) Which of the following is an accurate statement about high-constraint, laboratory research? A) It has high internal validity but may have low external validity. B) It is necessarily high in both external and internal validity. C) It is so constrained that it has no external validity. D) It must include at least four groups for proper controls. Answer: A Rationale: Option A correctly describes high-constraint laboratory research as having high internal validity but potentially lower external validity due to the controlled nature of laboratory settings. 19) Cook and Campbell (1979) noted that A) measures used in the field are often invalid. B) researchers cannot assume that findings from the lab will hold in the field. C) field research is not reliable and should not be used to answer questions of causality. D) All of the above Answer: B Rationale: Option B accurately reflects Cook and Campbell's observation that findings from laboratory research cannot be automatically assumed to generalize to field settings. 20) Since field settings are NOT exactly like the laboratory situation, laboratory findings cannot easily be assumed to have A) internal validity. B) construct validity. C) external validity. D) criterion-related validity. Answer: C Rationale: Option C correctly states that laboratory findings cannot be readily assumed to have external validity due to the differences between laboratory and field settings. 21) The high control and internal validity that are characteristic of experimental research in laboratories often mean a reduction in A) cost B) external validity. C) causal inferences D) all of the above Answer: B Rationale: Experimental research conducted in laboratories typically involves high control over variables, ensuring internal validity. However, this control often comes at the expense of external validity, as the findings may not generalize to real-world settings or diverse populations. 22) According to Cook and Campbell (1979), one of the disadvantages of using experimental control to infer causation is A) that it requires difficult and elaborate equipment. B) that we do not always want to learn about causation in controlled settings. C) that we do not always want to generalize our findings. D) that it requires the use of elaborate statistical techniques. Answer: B Rationale: Cook and Campbell (1979) noted that one of the disadvantages of relying solely on experimental control to infer causation is that researchers may not always want to learn about causation in controlled settings. This limitation highlights the importance of considering diverse contexts and real-world conditions in research. 23) If we conducted a highly controlled study that shows therapy X to be clearly superior to standard therapies Y and Z, what can we assume about the new therapy? A) Therapy X is superior under any and all conditions. B) Therapy X is superior under natural conditions. C) Therapy X is superior under laboratory conditions. D) Therapy X does not need field testing. Answer: C Rationale: If therapy X is shown to be superior to standard therapies Y and Z in a highly controlled study, it indicates superiority under laboratory conditions. However, this does not necessarily imply superiority under all conditions or in real-world settings without further field testing. 24) Which of the following is true for experiments in field settings? A) They cannot draw causal inferences. B) They can provide only observations of contingencies. C) They can allow causal inferences. D) They can generate hypotheses but not test them. Answer: C Rationale: Experiments conducted in field settings can allow for causal inferences, as they involve manipulating variables and observing their effects in real-world contexts. This enables researchers to draw conclusions about cause-and-effect relationships in natural environments. 25) What assumption is implicit in a Stop DWI (Driving While Intoxicated) program? A) The program will reduce the number of drunk drivers on the road. B) The program will reduce the number of highway fatalities. C) The program will reduce the amount of alcohol people drink. D) Both A and B Answer: D Rationale: The implicit assumption in a Stop DWI program is that implementing the program will lead to reductions in both the number of drunk drivers on the road and the number of highway fatalities, addressing the problem of driving while intoxicated. 26) Large scale social programs are generally enacted A) because they have been shown to be effective. B) for political reasons. C) after research shows that there is a social problem. D) All of the above Answer: B Rationale: Large-scale social programs are often enacted for political reasons, such as addressing public concerns or fulfilling campaign promises, rather than solely based on evidence of effectiveness. 27) Which of the following is NOT usually possible in field research? A) quasi-experimental design B) single-subjects design C) flexibility D) random assignment of participants Answer: D Rationale: Random assignment of participants is typically not feasible in field research due to the lack of control over environmental variables and the practical constraints of working in real-world settings. 28) An example of an appropriate and meaningful use of field research would be to A) test the effectiveness of social programs. B) provide a heuristic function. C) employ graduate students. D) establish causality. Answer: A Rationale: Field research can be appropriately used to test the effectiveness of social programs in real-world settings, providing valuable insights into their impact on individuals and communities. 29) One of the major problems with social programs is that A) they are well-meaning but useless. B) their supporters are seldom committed to testing the program's effectiveness. C) they are often unethical. D) they have nothing to offer psychology. Answer: B Rationale: One major problem with social programs is that their supporters may be less committed to rigorously testing their effectiveness, leading to implementation without sufficient evidence of their impact. 30) Psychologist Donald Campbell suggested that A) psychology should merge with sociology. B) psychology should stop using statistical analyses of data. C) developed countries should carry out social reforms in a controlled, experimental manner. D) research methods in psychology should not be concerned with external validity. Answer: C Rationale: Donald Campbell advocated for conducting social reforms in developed countries in a controlled, experimental manner, emphasizing the importance of experimental methods and causal inference in addressing social issues. 31) One of the goals of applied research is not only to understand phenomena, but to A) generate rival hypotheses. B) change the phenomena under study. C) impute causality. D) use the new understanding in some practical way. Answer: D Rationale: Applied research aims to not only understand phenomena but also to utilize the newfound understanding in practical applications. Option A may be a goal in basic research. Option B suggests an intervention, which may not always be the goal. Option C refers to a methodological aspect rather than a goal of applied research. 32) In doing research on a natural disaster, we would consider the natural event as A) the independent variable. B) an act of God. C) the dependent variable. D) something we can manipulate. Answer: A Rationale: In research on a natural disaster, the natural event itself is considered the independent variable, as it is the factor being studied for its effects on other variables. Option B is not a scientific designation. Option C is incorrect because the disaster event is not the outcome being measured. Option D is incorrect because natural disasters are not typically something researchers can manipulate. 33) An important reason for conducting field experiments is A) to take advantage of the potential for greater generalizability of the results over time. B) to enhance the internal validity of the study. C) to enhance the construct validity of the study. D) to take advantage of the potential for greater generalizability of results to other conditions. Answer: D Rationale: Field experiments are conducted to increase the generalizability of results to other conditions or settings beyond the laboratory, thereby enhancing external validity. Options A, B, and C are incorrect because they do not accurately describe the primary purpose of conducting field experiments. 34) Which of the following is NOT one of the three types of generalization distinguished in the text? Generalization of the results A) from the participants in the study to the larger population. B) of the study over time. C) from one study to another. D) from the conditions of the study to other conditions (i.e., generalization to other settings). Answer: C Rationale: Generalization from one study to another is not one of the three types distinguished in the text. Options A, B, and D accurately describe different types of generalization. 35) Which type of generalization may be enhanced by carrying out research in naturalistic settings? A) generalization of results from the participants in the study to the larger population B) generalization of the results of the study over time C) generalization of results from conditions of the study to other conditions (i.e. generalization to other settings) D) All of the above Answer: C Rationale: Conducting research in naturalistic settings may enhance generalization of results from the conditions of the study to other settings, as it allows for observations in real-world contexts. Options A and B do not specifically relate to naturalistic settings, and option D is incorrect because not all types of generalization are necessarily enhanced. 36) One of the problems in doing field research is that A) participants are not available. B) random assignment of participants to groups is often impossible. C) dependent measures cannot be taken. D) score data cannot be generated, and therefore, ANOVA cannot be used. Answer: B Rationale: Random assignment of participants to groups is often impossible in field research due to the lack of control over environmental conditions. Options A, C, and D are incorrect because they do not accurately describe typical problems associated with field research. 37) A major difficulty with drawing causal inferences in the field is A) the inability to obtain reliable measures. B) the inability to randomly assign participants. C) noncompliance of political leaders. D) laboratory equipment seems out of place in natural settings. Answer: B Rationale: The inability to randomly assign participants in field settings poses a major difficulty in drawing causal inferences, as it limits the researcher's ability to control for potential confounding variables. Options A, C, and D are incorrect because they do not accurately describe difficulties specific to drawing causal inferences in the field. 13.2 Quasi-Experimental Designs 1) Quasi-experimental designs are A) superior to experimental designs. B) similar to experimental designs but are not quite equal to them. C) often used in physics and chemistry. D) equal to experimental designs. Answer: B Rationale: Quasi-experimental designs share similarities with experimental designs but lack certain elements, such as random assignment to conditions, making them not quite equal to experimental designs. Options A, C, and D are incorrect because they do not accurately describe quasi-experimental designs. 2) The use of quasi-experimental designs is recommended A) only in situations in which full experimental controls can be used. B) in situations in which there is no causal question to be addressed. C) in situations in which full experimental controls cannot be employed, but where it is important to address a causal question. D) only in laboratory settings. Answer: C Rationale: Quasi-experimental designs are recommended in situations where full experimental controls cannot be employed but where it is important to address a causal question. Options A, B, and D do not accurately describe the recommended use of quasi-experimental designs. 3) Quasi-experimental designs are different from low-constraint methods in that A) quasi-experimental methods attempt to answer questions of causality. B) low-constraint methods attempt to answer questions of causality. C) unlike low-constraint methods, quasi-experiments always directly manipulate the independent variable. D) unlike low-constraint methods, quasi-experiments usually assign participants randomly to groups. Answer: A Rationale: Quasi-experimental designs attempt to answer questions of causality, whereas low-constraint methods may not necessarily aim to establish causal relationships. Options B, C, and D incorrectly describe differences between quasi-experimental designs and low-constraint methods. 4) What term is used for designs that are like experimental designs but NOT quite equal to them and from which we can still draw causal inferences? A) correlational designs B) differential designs C) quasi-experimental designs D) naturalistic-experimental designs Answer: C Rationale: Quasi-experimental designs are similar to experimental designs but lack full experimental control, yet they still allow for drawing causal inferences, making option C the correct answer. 5) Which of the following statements do NOT describe quasi-experimental designs? A) They are like experimental design but are not quite equal to them. B) They have a causal hypothesis. C) They have only one level of the independent variable. D) They have some type of manipulation in order to compare two or more conditions. Answer: C Rationale: Quasi-experimental designs typically involve manipulation of the independent variable to compare different conditions, so option C does not accurately describe them. 6) How are quasi-experimental designs different from low-constraint methods? A) Quasi-experiments have only one level of the independent variable. B) Low-constraint methods do not attempt to answer questions of causality. C) Low-constraint methods try to control confounding variables. D) Quasi-experiments take place in a laboratory. Answer: B Rationale: Quasi-experimental designs still aim to draw causal inferences, unlike low-constraint methods, which do not focus on questions of causality, making option B the correct answer. 7) Using quasi-experiments to draw causal inferences in field experiments A) may give us less than high confidence but will still give us more information than not experimenting at all. B) is more costly and more time consuming than answering the same question in the laboratory. C) is based on new statistical procedures. D) All of the above Answer: A Rationale: Quasi-experiments in field settings may provide less confidence in causal inferences compared to laboratory experiments but still offer valuable information, making option A the correct answer. 8) Quasi-experiments are unlike experiments in that A) they do not always include a comparison of at least two levels of an independent variable. B) actual manipulation of the independent variable is not always clearly defined or under the experimenter's control. C) often the experimenter cannot assign participants to groups in an unbiased manner. D) Both B and C Answer: D Rationale: Quasi-experiments may lack clear manipulation of the independent variable and unbiased assignment of participants, distinguishing them from traditional experiments, as stated in options B and C. 9) Quasi-experimental designs attempt to answer A) questions of causality. B) low-constraint questions. C) manipulated events questions. D) laboratory manipulation questions. Answer: A Rationale: Quasi-experimental designs aim to draw causal inferences, making option A the correct answer. 10) Which of the following is NOT a characteristic of quasi-experimental designs? A) Causal hypotheses are stated. B) The participants are usually assigned to groups in an unbiased manner. C) At least two levels of the independent variable are included. D) Unlike low-constraint methods, quasi-experiments usually assign participants to groups. Answer: B Rationale: Quasi-experimental designs often lack unbiased assignment of participants, unlike traditional experiments, as indicated in option B. 11) Which of the following is NOT a characteristic of quasi-experimental designs? A) The researcher includes specific procedures for testing hypotheses. B) The researcher usually must accept already existing groups. C) The researcher always directly manipulates the independent variable. D) The researcher includes controls for threats to validity. Answer: C Rationale: Quasi-experimental designs may not always involve direct manipulation of the independent variable, unlike traditional experimental designs, making option C the correct answer. 12) Two basic quasi-experimental designs mentioned in the text are A) the nonequivalent control-group design and the differential design. B) the repeated-measures design and the nonequivalent control-group design. C) the interrupted time-series design and the case-study design. D) the nonequivalent control-group design and the interrupted time-series design. Answer: D Rationale: The nonequivalent control-group design and the interrupted time-series design are commonly used quasi-experimental designs, as stated in option D. 13) Which of the following is NOT a characteristic of quasi-experimental designs? A) They include specific procedures for testing hypotheses. B) They include some controls for threats to validity. C) They include random assignment of participants to groups. D) They include a causal hypothesis. Answer: C Rationale: Quasi-experimental designs often lack random assignment of participants to groups, unlike traditional experimental designs, making option C the correct answer. 14) Using already existing groups for research implies that A) participants cannot be randomly assigned. B) the research project is doomed to fail. C) any comparisons made between the groups are meaningless. D) participants have already been randomly assigned. Answer: A Rationale: When using already existing groups for research, participants cannot be randomly assigned to different conditions or groups. This lack of random assignment means that the researcher must work with groups that already exist in the population, limiting the control over potential confounding variables. 15) When naturally existing groups can be shown to be similar to each other on most relevant variables, an appropriate quasi-experimental design would be the A) equivalent control-group design. B) nonequivalent control-group design. C) quasi-equivalent control-group design. D) preexisting control-group design. Answer: B Rationale: The nonequivalent control-group design is appropriate when naturally existing groups are used, and they can be shown to be similar on most relevant variables. This design helps to control for potential confounding variables and allows for comparisons between groups. 16) What is an example of a quasi-experimental design? A) nonequivalent control-group design B) interrupted time-series design C) ABA reversal design D) Both A and B Answer: D Rationale: Quasi-experimental designs include designs like the nonequivalent control-group design and interrupted time-series design. These designs lack random assignment but still allow researchers to study the effects of an independent variable on a dependent variable. 17) Why is it best when testing a causal hypotheses to compare groups that are randomly assigned? A) It is more likely that the groups are equivalent at the beginning of the study. B) Random assignment assures that the participants are unaware of what is being studied. C) Random assignment assures that the experimenter is blind to the condition of the participant. D) All of the above Answer: A Rationale: Random assignment helps ensure that the groups are equivalent at the beginning of the study, reducing the likelihood of systematic differences between groups that could confound the results. Options B and C refer to blinding procedures, which are important but not directly related to the advantages of random assignment. 18) If groups are the same on the dependent variable at the start of the study, A) they are said to be equivalent. B) the same type of conclusions can be drawn as if they were randomly assigned. C) they may still differ on variables that could be potentially confounding. D) Both A and B Answer: C Rationale: Even if groups are the same on the dependent variable at the start of the study, they may still differ on other variables that could potentially confound the results. Therefore, conclusions drawn from such groups may not be as strong as those drawn from randomly assigned groups. 19) The best alternative to using groups that have been randomly assigned to conditions is to use A) groups that have been assigned in a biased manner. B) a no control-group design. C) a nonequivalent control-group design in which the groups appear similar on relevant variables. D) a nonequivalent control-group design in which the groups appear different on relevant variables. Answer: C Rationale: When random assignment is not possible, the best alternative is to use a nonequivalent control-group design where efforts are made to ensure that the groups appear similar on relevant variables. This helps to control for potential confounding variables and strengthens causal inferences. 20) One of the most serious problems with nonequivalent control-group designs is that A) groups may be different at the outset on the dependent measures. B) groups may be different at the outset on the key independent variables. C) difference scores are cumbersome to use in statistical analyses. D) participant recruitment is much more difficult than it is in other types of designs. Answer: A Rationale: One of the most serious problems with nonequivalent control-group designs is that groups may be different at the outset on the dependent measures. This threatens the internal validity of the study because it becomes difficult to attribute any observed differences between groups to the manipulation of the independent variable. 21) In the nonequivalent control-group design, A) the more similar the natural groups are to each other, the more confidence we can have in causal conclusions. B) the more dissimilar the natural groups are, the more confidence we can have in causal conclusions. C) the natural groups must be tested to assure equivalence on the dependent measure. D) the natural groups must be tested to assure equivalence on the independent variable. Answer: A Rationale: In the nonequivalent control-group design, the more similar the natural groups are to each other, the more confidence we can have in causal conclusions. This similarity reduces the likelihood that observed differences between groups are due to preexisting differences rather than the experimental manipulation. 22) Which is an example of a study in which a nonequivalent control-group design could be used? A) We want to evaluate the effectiveness of an anti-smoking campaign using two different schools. We suspect that one school has a higher rate of smoking than the other. B) A new treatment approach for hyperactivity is to be tested in a special school using three different classes. The classes, however, are at different levels of hyperactivity. C) A company wants to test their employees' attitudes toward work as they may be affected by new work rules. One department is selected as an experimental group, and the other is designated as the control group. However, there is evidence that the first department has more positive work attitudes than the second. D) All of the above Answer: D Rationale: All of the scenarios described involve situations where random assignment may not be feasible, and a nonequivalent control-group design could be used to compare groups that already exist in the population. 23) What is a major problem with nonequivalent groups? A) The groups may differ on the dependent measure(s) at the start of the study. B) There may be confounding variables not controlled by random assignment. C) The groups may differ on the independent measure at the start of the study. D) Both A and B Answer: D Rationale: A major problem with nonequivalent groups is that they may differ on both the dependent measure(s) and potentially confounding variables at the start of the study, threatening the internal validity of the research findings. Therefore, both options A and B are correct. 24) According to Cook and Campbell (1979), if you have evidence that the naturally-occurring groups in a nonequivalent control-group design are NOT equivalent on potential confounding variables, A) you should find new naturally-occurring groups that are equivalent. B) you may be able to draw conclusions if you carefully evaluate all potential threats to validity. C) you should continue with the experiment as planned, because the groups only need to be equivalent on the dependent measure. D) you may be able to draw conclusions if you reassign participants to different groups. Answer: B Rationale: Cook and Campbell suggest that even if the groups are not equivalent, careful consideration and evaluation of potential threats to validity may still allow for meaningful conclusions to be drawn from the study. 25) A potential problem facing nonequivalent control-group designs is that A) unequal group size may hinder the analyses. B) these designs are unethical. C) groups may differ on variables other than the dependent variable. D) these designs work better with animals than with humans. Answer: C Rationale: Nonequivalent control-group designs may encounter the problem that the groups being compared differ on variables other than the dependent variable, potentially confounding the interpretation of results. 26) The two chief problems in nonequivalent control-group designs are related to the fact that A) since the groups exist prior to the study, participants are assigned to groups in a random manner. B) participants are assigned to groups in a unbiased manner. C) there is no experimental manipulation. D) since the groups exist prior to the study, participants cannot be assigned in an unbiased manner. Answer: D Rationale: One of the chief problems in nonequivalent control-group designs is that participants cannot be assigned to groups in an unbiased manner, as the groups already exist prior to the study, which can lead to issues with internal validity. 27) In which of the following is comparing a pretest-posttest scores for experimental and control groups a typical strategy? A) all experimental studies B) all quasi-experimental studies C) nonequivalent control-group designs D) between-subjects designs Answer: C Rationale: Comparing pretest-posttest scores for experimental and control groups is a typical strategy in nonequivalent control-group designs to assess the impact of an intervention or treatment. 28) Two common confounding variables that must be controlled in a nonequivalent control-group design are A) selection and regression to the mean. B) selection and maturation. C) maturation and regression to the mean. D) sequence effects and maturation. Answer: A Rationale: Selection (systematic differences between groups at baseline) and regression to the mean (extreme scores tend to regress toward the mean upon retesting) are common confounding variables in nonequivalent control-group designs that need to be controlled to ensure internal validity. 29) In a nonequivalent control-group design, if the experimental and control groups differ on the dependent measure at pretest, we must rule out confounding due to A) regression to the mean. B) instrumentation. C) selection. D) Both A and C Answer: D Rationale: If the experimental and control groups differ on the dependent measure at pretest, confounding due to regression to the mean (A) and selection (C) must be ruled out to ensure that observed differences are not due to these factors. 30) In a nonequivalent control-group design, if the experimental and control groups are equivalent on the dependent measure at pretest, concern about confounding due to ________ would be reduced. A) carry-over effects B) selection C) instrumentation D) history Answer: B Rationale: If the experimental and control groups are equivalent on the dependent measure at pretest, concerns about confounding due to selection are reduced, as it suggests that the groups were comparable at the outset. 31) In a nonequivalent control-group design, if the experimental and control groups are equivalent at pretest and show equivalent change over time, you might conclude that A) there is no effect of the independent variable. B) there is no effect of the dependent variable. C) the independent variable had an effect only on the experimental group. D) the independent variable had an effect only on the control group. Answer: A Rationale: If both groups are equivalent at pretest and show equivalent change over time, it suggests that the treatment had no effect, as both groups changed similarly regardless of treatment condition. 32) In a nonequivalent control-group design, if the experimental and control groups are equivalent at pretest and the experimental group shows a large change while the control group does NOT change from pretest to posttest, you might conclude that A) there is an effect of the independent variable. B) there is an effect of the dependent variable. C) the changes are due to selection. D) Both A and C Answer: A Rationale: If the experimental group shows a significant change while the control group does not, it suggests that the independent variable (treatment) had an effect on the experimental group's outcome, supporting the conclusion of an effect. 33) In a nonequivalent control-group design, if neither the experimental nor the control group change over time, you might conclude that A) there is an effect of the independent variable. B) the manipulation of the independent variable had no effect. C) the groups were not equivalent at pretest. D) the effect is due to regression to the mean. Answer: B Rationale: If neither group changes over time, it suggests that the treatment had no effect on the outcome variable, indicating that the manipulation of the independent variable had no effect. 34) In a nonequivalent control-group design, if the experimental and control groups are NOT equivalent at pretest and both groups change in the same manner (e.g., both groups increase an equal amount on the measure), you might conclude that A) the manipulation of the independent variable affected both groups. B) the effect was due to selection. C) a maturation process or historical event common to the two groups caused the change in scores. D) Both A and B Answer: C Rationale: If both groups change in the same manner despite not being equivalent at pretest, it suggests that a common factor, such as maturation or a historical event, may have influenced the change in scores rather than the manipulation of the independent variable. 35) In a nonequivalent control-group design with groups that differ at pretest, if the control group shows a small change while the experimental group shows a marked change over time, you might conclude that A) there is no effect of the independent variable. B) there may have been an effect of the independent variable, but it may have been confounded by regression to the mean. C) the effect is due to instrumentation. D) the effect is due to demand characteristics of the setting. Answer: B Rationale: If the experimental group shows a marked change while the control group shows a small change, it suggests that there may have been an effect of the independent variable. However, this effect may be confounded by regression to the mean, where extreme scores at pretest tend to move closer to the mean at posttest. 36) Giving the dependent measure before any manipulation of the independent variable in a nonequivalent control-group design allows the researcher to A) assess the similarity of the experimental and control groups. B) determine what the range of scores will be after the manipulation. C) drop any participants that differ significantly from the norm. D) All of the above Answer: A Rationale: Administering the dependent measure before any manipulation allows the researcher to assess the similarity of the experimental and control groups at baseline, which is crucial for interpreting the results of the study. 37) It is important to assess the similarity of the experimental and control groups before any manipulation of the independent variable in a nonequivalent control-group design because A) if the groups are significantly different, direction of causality cannot be determined. B) the more similar the groups are, the greater control we have. C) as the similarity between groups at the beginning of the study increases, the design approaches experimental designs. D) Both B and C Answer: D Rationale: Assessing the similarity of the experimental and control groups before manipulation allows for greater control over potential confounding variables and increases the internal validity of the study, bringing it closer to experimental designs. 38) If you suspect that the groups in a nonequivalent control-group design differ on variables other than the dependent variable, you should A) systematically identify potential confounding variables. B) measure the variables that may be confounded. C) carefully rule out the alternative hypotheses. D) All of the above Answer: D Rationale: When suspecting group differences on variables other than the dependent variable, it's important to systematically identify potential confounding variables, measure them, and carefully rule out alternative hypotheses to ensure the validity of the study. 39) Causality is inferred in a nonequivalent control-group design when A) the control group remains essentially the same from pretreatment to post-treatment measures but the experimental group changes markedly in the predicted direction. B) the experimental group remains essentially the same from pretreatment to post-treatment measures but the control group changes markedly in the predicted direction. C) the control group and the experimental group remain essentially the same from pretreatment to post-treatment measures. D) the control group and the experimental group change markedly in the predicted direction. Answer: A Rationale: In a nonequivalent control-group design, causality is inferred when the experimental group changes markedly in the predicted direction while the control group remains essentially the same from pretest to post-test, indicating that the manipulation of the independent variable influenced the observed changes. 40) In one possible outcome of a nonequivalent control-group design, the control group did NOT change but the experimental group, which initially scored lower than the control group, scored much higher than the control group at posttest. This is called a A) confound due to selection. B) crossover effect. C) criss-cross effect. D) serendipitous finding. Answer: B Rationale: In a crossover effect, the experimental group initially scoring lower than the control group shows a marked improvement and scores much higher than the control group at posttest, indicating a reversal of initial group differences. 41) In a crossover effect, maturation is an unlikely alternative hypothesis because A) maturation cannot affect the control group. B) maturation cannot affect the experimental group. C) if maturation were responsible, it is unlikely that the effect would markedly differ for both groups. D) it is likely that both groups mature at a different rate. Answer: C Rationale: Maturation is an unlikely alternative hypothesis in a crossover effect because if maturation were responsible, it is unlikely that the effect would markedly differ for both groups, as observed in a crossover effect. 42) In a crossover effect, regression to the mean is an unlikely alternative hypothesis because A) regression to the mean does not operate in quasi-experimental designs. B) participants are not selected on the basis of extreme scores in nonequivalent control-group designs. C) a crossover effect occurs prior to any statistical analyses. D) the experimental group increased, not only to the mean of the control group, but changed beyond it. Answer: D Rationale: Regression to the mean is an unlikely alternative hypothesis in a crossover effect because the experimental group not only increased to the mean of the control group but changed beyond it, indicating a genuine effect rather than regression to the mean. 43) In a nonequivalent control-group design, if the control group does NOT change but the experimental group changes markedly, even going far beyond the level of the control group, you might A) have considerable confidence in a causal inference. B) conclude that there is an effect of the independent variable. C) conclude that there is a crossover effect. D) All of the above Answer: D Rationale: If the control group does not change but the experimental group changes markedly, it provides evidence for a causal inference regarding the effect of the independent variable, and it may also indicate a crossover effect where the experimental group surpasses the control group's level. 44) In a nonequivalent control-group design, if the control group does NOT change but the experimental group changes markedly, even going far beyond the level of the control group, it is called a A) crossover effect. B) crossmark effect. C) experimental-control effect. D) hybrid effect. Answer: A Rationale: A crossover effect occurs when the experimental group changes markedly, surpassing the level of the control group, in a nonequivalent control-group design. This phenomenon can lead to misinterpretation of results if not adequately addressed. 45) Regression to the mean can be at least partially controlled in the single-group situation by A) using an interrupted time-series design. B) changing the measure used at pretest and posttest. C) using a quasi-experimental design. D) using a pretest-posttest design. Answer: A Rationale: Interrupted time-series designs involve measuring a single group multiple times before and after some manipulation. By including multiple measures over time, researchers can better assess whether changes in the dependent variable are due to the manipulation or simply to regression to the mean. 46) A type of design in which a single group of participants is measured several times both before and after some manipulation is termed A) an interrupted time-series design. B) an interrupted manipulation design. C) a nonequivalent control-group design. D) a longitudinal design. Answer: A Rationale: Interrupted time-series designs involve measuring a single group of participants multiple times before and after a manipulation or intervention. This allows researchers to assess changes in the dependent variable over time and determine the effects of the manipulation. 47) Time-series designs are variations of A) naturalistic designs. B) within-subjects designs. C) between-subjects designs. D) management designs. Answer: B Rationale: Time-series designs involve measuring the same group of participants repeatedly over time, making them variations of within-subjects designs because each participant serves as their own control across different time points. 48) In its general form, the time-series design is similar to a A) between-subjects design. B) nonequivalent control-group design. C) simple pretest-posttest design. D) main effects model. Answer: C Rationale: In its general form, the time-series design resembles a simple pretest-posttest design, where the dependent variable is measured before and after some manipulation or intervention for a single group of participants. 49) One of the major potential confounding factors in a simple pretest-posttest study is A) placebo effects. B) frustration effects. C) effects due to the treatment. D) regression to the mean. Answer: D Rationale: Regression to the mean is a major potential confounding factor in simple pretest-posttest studies, where extreme scores at pretest are likely to regress toward the mean at posttest, regardless of any treatment effects. 50) Baseline measures are taken A) before any manipulations are made. B) immediately following the manipulation. C) only on the experimental group. D) only on the control group. Answer: A Rationale: Baseline measures are taken before any manipulations or interventions are introduced, serving as a reference point to assess changes in the dependent variable over time or in response to a manipulation. 51) What is the term used for the design in which a single group of participants is measured several times both before and after an event or manipulation? A) single-subject design B) general quasi-experimental design C) between-subjects design D) interrupted time-series design Answer: D Rationale: An interrupted time-series design involves measuring a single group of participants multiple times before and after an event or manipulation, allowing researchers to assess changes in the dependent variable over time. 52) In interrupted time-series designs, A) the experiences of a single participant are manipulated and recorded. B) two or more groups of participants are measured several times. C) a single group of participants is measured several times both before and after some event or manipulation. D) a single group of participants is measured once at pretreatment and once at post-treatment. Answer: C Rationale: Interrupted time-series designs involve measuring a single group of participants multiple times both before and after some event or manipulation, allowing researchers to assess changes in the dependent variable over time. 53) What is the difference between a simple pretest-posttest design and an interrupted time-series design? A) The time-series design uses multiple measures taken at several points in time both before and after the manipulation. B) The simple pretest-posttest design uses a single measure at pretreatment and post-treatment. C) The time-series uses a series of measures that are "interrupted" by the manipulation. D) All of the above Answer: D Rationale: The difference between a simple pretest-posttest design and an interrupted time-series design includes the use of multiple measures over time and the "interrupted" nature of the time-series design, which captures changes in the dependent variable before and after the manipulation. 54) In an interrupted, time-series design, the time-series of measures are "interrupted" by A) a change in the time between measures. B) a change in measures. C) a manipulation by the researcher. D) the time of day. Answer: C Rationale: In an interrupted, time-series design, the time-series of measures are "interrupted" by a manipulation by the researcher. This manipulation could be the introduction or removal of a treatment, intervention, policy change, or any other deliberate alteration made by the researcher during the course of the study. 55) What is a major potential confounding factor in the simple pretest-posttest design that is controlled for in the interrupted time-series design? A) instrumentation B) regression to the mean C) selection D) maturation Answer: B Rationale: Regression to the mean is a major potential confounding factor in the simple pretest-posttest design. In an interrupted time-series design, this factor is controlled for because the design involves multiple measurements over time, allowing researchers to observe trends and changes in the dependent variable beyond mere chance fluctuations. 56) Which of the following is similar to a pretest-posttest design, but with more measures? A) ex post facto design B) general quasi-experimental designs C) program evaluation designs D) interrupted time-series design Answer: D Rationale: Interrupted time-series design is similar to a pretest-posttest design but involves multiple measures taken over time, allowing for a more detailed examination of changes in the dependent variable before, during, and after an intervention or manipulation. 57) Regression to the mean is a major potential confounding factor in which of the following? A) posttest-only designs B) pretest-posttest designs C) all factorial designs D) within-subjects designs Answer: B Rationale: Regression to the mean is a major potential confounding factor in pretest-posttest designs. It refers to the tendency for extreme scores on a pretest to move closer to the mean on subsequent measurements, regardless of any intervention or treatment administered. This can lead to erroneous conclusions if not properly controlled for. 58) The interrupted time-series design is able to control for regression to the mean because A) by employing multiple measures, it gives several points of comparison over a long period of time. B) it adds a constant to all scores. C) it multiplies all scores by a constant. D) by taking refined measurements, the design uses only one composite score, thereby reducing error variance. Answer: A Rationale: The interrupted time-series design controls for regression to the mean by employing multiple measures over time, allowing researchers to observe trends and patterns beyond mere chance fluctuations. This reduces the likelihood of attributing changes solely to regression to the mean. 59) One potential confounding factor in the simple pretest-posttest design, which the interrupted time-series designs controls for, is A) regression to the mean. B) experimenter bias. C) the double-blind effect. D) an order effect. Answer: A Rationale: One potential confounding factor in the simple pretest-posttest design is regression to the mean. Interrupted time-series designs control for this factor by incorporating multiple measurements over time, providing a more comprehensive understanding of changes in the dependent variable. 60) Which of the following is an example of an interrupted time-series design? A) A school wants to try a new teaching program and test its effectiveness. Mrs. Jones' first grade class is given the new program. Mrs. Jones' class and Mrs. Brown's first grade class (who received the traditional teaching method) both take an exam at the end of the year designed to determine the amount that each class learned. B) The state reduces the speed limit from 65 to 55. Records are gathered showing the number of highway deaths per month for the year prior to the reduction in the speed limit and for the year following the reduction. C) A new drug treatment for AIDS is being tested. Patients on a waiting list are randomly assigned to receive either the new drug treatment or a placebo. The number of visits to the doctor that the drug group and the placebo group make during the next year will be monitored and compared. D) Victims of a hurricane are interviewed and asked their opinions about the needs that were not met following the disaster. This information is then turned into a social program to aid disaster victims. Answer: B Rationale: Option B describes an interrupted time-series design where the time-series of measures (number of highway deaths per month) is interrupted by a change in policy (reduction in speed limit). This interruption allows researchers to assess the impact of the policy change on the outcome variable (highway deaths). 61) Which of the following is NOT mentioned in the text as a feature of interrupted time-series designs? A) They can be used on large-scale or small-scale events. B) They can be used to evaluate events that have already occurred. C) They can use data already collected. D) They are inflexible designs. Answer: D Rationale: Interrupted time-series designs are flexible designs that can be used in various situations and settings, ranging from large-scale events to small-scale interventions. They can evaluate events that have already occurred and utilize data already collected, making them versatile tools for studying the effects of interventions or policy changes over time. 62) The interrupted time-series design A) is a flexible design that can be used in many situations. B) can use data that already have been gathered. C) can be used to evaluate events that have already occurred. D) All of the above. Answer: D Rationale: All of the statements are true regarding interrupted time-series designs. They are flexible designs that can be adapted to different situations, they can utilize pre-existing data, and they can evaluate events that have already occurred, making them valuable tools for studying the effects of interventions or policy changes. 63) Interrupted time-series designs are most useful in providing a good way of testing causal hypotheses under A) very controlled laboratory conditions. B) many field conditions. C) anesthesia. D) the usual constraints of experimentation. Answer: B Rationale: Interrupted time-series designs are particularly useful in field conditions where researchers cannot exert as much control as in laboratory settings. They allow for the evaluation of real-world interventions or policy changes and provide a robust method for testing causal hypotheses in applied settings. 64) If we wanted to determine if hospital admissions for asthmatic attacks were related to air quality, which of the following would be the most appropriate design? A) ex post facto B) pretest-posttest C) interrupted time-series D) independent groups Answer: C Rationale: Interrupted time-series designs are particularly suited for studying the effects of interventions or events that occur over time, making them appropriate for assessing the relationship between changes in air quality and hospital admissions for asthmatic attacks. 65) By using an interrupted time-series design with a single group of participants, researchers can A) randomly assign participants to treatment conditions. B) take advantage of data already gathered over a long period of time. C) create a simulated between-subjects design. D) attenuate reliability. Answer: B Rationale: Interrupted time-series designs with a single group allow researchers to utilize existing data gathered over time, providing valuable information about the effects of interventions or events without the need for a control group. 66) Interrupted time-series designs are particularly useful in all of the following situations EXCEPT A) large-scale studies. B) evaluating events that have already occurred. C) clinical and naturalistic settings. D) situations in which it is impossible to make more than one measurement. Answer: D Rationale: Interrupted time-series designs are not suitable for situations where it is impossible to make more than one measurement. These designs rely on multiple measurements before and after the intervention or event to assess its effects over time. 67) The two most important potential confounding factors in time-series designs are A) history and instrumentation B) selection and regression to the mean C) regression to the mean and cohort effects D) cohort effects and history Answer: A Rationale: History and instrumentation are the two most important potential confounding factors in time-series designs. History refers to external events or factors that occur during the study period and may influence the outcome, while instrumentation concerns changes in measurement procedures or instruments over time. 68) If the state of Utah were to use an interrupted time-series design to study the effects of raising its speed limit to 65 mph, what type of confounding variables might present a problem? A) Confounding due to selection and cohort effects. B) Confounding due to history and instrumentation. C) Confounding due to maturation and testing. D) Confounding due to regression to the mean. Answer: B Rationale: In this scenario, confounding due to history (e.g., changes in traffic patterns or enforcement) and instrumentation (e.g., changes in speed measurement technology) might present problems because they could influence the observed effects of raising the speed limit. 69) Any procedure that requires measurement over a fairly long period of time is particularly vulnerable to confounding by A) entropy B) selection ratio C) geography D) history. Answer: D Rationale: Confounding by history is particularly problematic in procedures that require measurement over a long period of time because external events or factors occurring during the study period may influence the outcome, leading to confounding. 70) Which of the following confounding factors is most likely to affect time-series designs? A) participant selection B) regression to the mean C) maturation D) history Answer: D Rationale: History, or external events or factors occurring during the study period, is the most likely confounding factor to affect time-series designs, as it may influence the observed outcomes over time. 71) Under what conditions is history a potential confounding factor? A) When the pretest measures are at an unusually high level. B) When a disproportionate number of participants are lost from one condition. C) When participant selection has been biased. D) When the study takes a fairly long time to complete. Answer: D Rationale: History is a potential confounding factor when the study takes a fairly long time to complete because external events or factors occurring during the study period may influence the outcomes being measured. 72) History is a potential confounding factor in the interrupted time-series design because A) the time-series design takes a fairly long period of time. B) all of the participants are exposed to the same conditions. C) the participants are all considered one group despite individual differences in experience. D) the momentum of multiple measures is disrupted by the manipulation. Answer: A Rationale: History is a potential confounding factor in interrupted time-series designs because these designs typically involve measurement over a long period of time, increasing the likelihood that external events or factors (i.e., history) may influence the observed outcomes. 73) An example of history confounding the results of an interrupted time-series design on the effect of a speed limit reduction could be A) an increased number of patrol cars. B) an increased number of tickets being given. C) an increase in the fine for tickets. D) All of the above Answer: D Rationale: All of the provided examples could represent instances of history confounding the results of an interrupted time-series design on the effect of a speed limit reduction. Each represents an external event or factor that could influence the observed outcomes over time. 74) In time-series designs, what are the two major potential confounding factors that are difficult to control without the addition of a comparison group? A) instrumentation and history B) regression to the mean and instrumentation C) history and attrition D) instrumentation and selection Answer: A Rationale: Instrumentation refers to changes in the measuring procedures over time, while history refers to external events or factors that could influence the observed outcomes. Both instrumentation and history can confound the interpretation of results in time-series designs if not controlled for, making them major potential confounding factors. 75) Two potentially confounding factors in the interrupted time-series design are A) maturation and testing effects. B) regression to the mean and selection. C) history and instrumentation. D) maturation and regression to the mean. Answer: C Rationale: In interrupted time-series designs, history (external events or factors) and instrumentation (changes in measuring procedures) are two potential confounding factors that can affect the observed outcomes if not adequately controlled for. 76) An interrupted time-series design is particularly useful when A) there are no issues of causality to be addressed. B) the presumed causal event occurs to only some members of a population. C) the presumed causal event occurs to all members of a population. D) the research is concerned with attempting to predict aggressive behavior in children. Answer: C Rationale: Interrupted time-series designs are particularly useful when the presumed causal event, such as a policy change or intervention, occurs to all members of a population, allowing researchers to observe the effects of the event over time. 77) If researchers were interested in studying the effects of New York State's raising of the drinking age to 21 on traumatic brain injury, what type of research design would be ideal? A) a standard 2 X 2 factorial B) an equivalent control-group design C) a between-subjects design D) an interrupted time-series design Answer: D Rationale: An interrupted time-series design would be ideal for studying the effects of a policy change (raising the drinking age) on a particular outcome (traumatic brain injury) over time, allowing researchers to observe trends and changes in the outcome before and after the intervention. 78) In a study of the effects of a reduced speed limit on arrests for speeding, it was discovered that the police had switched from radar to Vascar (a new way of measuring the speed of an oncoming car). What possible confounding factor did this switch introduce? A) instrumentation B) history C) maturation D) regression to the mean Answer: A Rationale: The switch from radar to Vascar introduces a potential confounding factor of instrumentation, as it involves changes in the measuring procedures over time, which could affect the observed outcomes (arrests for speeding). 79) In order for a change in a time graph to be interpreted as significant in a time-series study, the change must A) not show up very frequently on the graph. B) be quite sharp. C) be slight or gradual. D) persist for a long time. Answer: B Rationale: A significant change in a time graph in a time-series study is typically interpreted as being quite sharp, indicating a clear deviation from the baseline or expected pattern. 80) In interpreting a time graph, a post-intervention change that appears ________ is difficult to interpret. A) on the graph B) steep or sharp C) slight or gradual D) None of the above Answer: C Rationale: A post-intervention change that appears slight or gradual on a time graph is difficult to interpret because it may not clearly indicate a significant deviation from the baseline or expected pattern, making it challenging to determine the effects of the intervention. 81) What is the term used for the potential confounding variable in a time-series study in which the measuring procedures may be inadvertently changed over time? A) measurement error B) history C) instrumentation D) attrition Answer: C Rationale: Instrumentation refers to changes in the measuring procedures over time, which can introduce bias or confound the interpretation of results in a time-series study. 82) Why is instrumentation a potential threat to validity in a time-series design? A) When participants take the same measures over and over again, they tend to get "better" at them. B) When people initiate new approaches or programs, there also may be accompanying changes in the way records are made and kept. C) The more times something is measured, the less reliable it is. D) When a study is carried out over a long period of time, participants tend to drop out or leave. Answer: B Rationale: Instrumentation can threaten validity in a time-series design because changes in the measuring procedures over time can affect the observed outcomes, introducing bias or confounding variables. 83) In an interrupted times-series study on the effect of a reduced speed limit, the change in the time graph must A) be a sharp drop followed by a gradual increase back to the baseline to be interpreted as anything other than normal fluctuation. B) be slight or gradual at the time of the "interruption" to be interpreted as anything but normal fluctuation. C) be quite sharp at the time of the "interruption" to be interpreted as anything other than normal fluctuation. D) be a sharp increase followed by a sharp decrease to be interpreted as anything but normal fluctuation. Answer: C Rationale: In an interrupted time-series study, a significant change in the time graph, such as a policy change or intervention, is typically interpreted as being quite sharp at the time of the interruption, indicating a clear deviation from the baseline or expected pattern. 84) In time-series designs, comparison groups A) are not possible to include. B) are helpful. C) always include random assignment. D) cannot exceed two groups. Answer: B Rationale: Comparison groups are helpful in time-series designs because they provide a basis for comparison to assess the impact of an intervention or treatment over time. They allow researchers to control for confounding variables and better understand the effects of the intervention. 85) The power of an interrupted time-series design can be increased by A) adding one or more comparison groups. B) studying important legal and regulatory phenomena, such as child car seats. C) random assignment of participants to groups. D) asking participants to practice regression to the mean. Answer: A Rationale: Adding one or more comparison groups to an interrupted time-series design can increase its power by providing additional data points for comparison, which strengthens the ability to draw conclusions about the effects of the intervention. 86) By adding comparison groups in a study of the effects of child seat belt legislation in California, Guerin and MacKinnon (1985) were able to A) perform more high-powered statistical procedures. B) attenuate the reliability of their observations. C) study the effects of other types of legislation on their target group. D) increase confidence in their conclusions. Answer: D Rationale: Adding comparison groups in the study increased confidence in the conclusions drawn from the research by providing a basis for comparison and controlling for potential confounding variables, thus strengthening the validity of the findings. 87) The study conducted in California to study the effects of child passenger restraint law for children under age 4 included two control groups. What were the two groups? A) Children over age 4 and children under age 4 in another state. B) Adults with children and children over age 4 in another state. C) Adults in California and adults in another state. D) Children with car seats and children without car seats. Answer: A Rationale: The two control groups included children over age 4 and children under age 4 in another state. This setup allowed researchers to compare the effects of the legislation in California to a similar population in another jurisdiction without the legislation. 88) The two control groups in the California study on the effect of a new child passenger restraint law were added to control for A) instrumentation. B) history. C) maturation. D) Both A and B Answer: D Rationale: The two control groups were added to control for potential confounding variables such as instrumentation (changes in measurement methods) and history (external events that could influence outcomes), hence addressing both factors A and B. 89) Addressing the null hypothesis in time-series designs is done by A) simply inspecting the time graph B) using simple t-tests. C) sophisticated statistical procedures D) computing the correlation coefficient. Answer: C Rationale: Time-series designs often require sophisticated statistical procedures to address the null hypothesis, such as autoregressive integrated moving average (ARIMA) modeling or interrupted time-series analysis, due to the complex nature of the data and the need to account for temporal dependencies and potential confounding variables. 13.3 Program Evaluation Research 1) Program evaluation research is one of the fastest growing areas of psychological research because A) research that encompasses more than one study is more cost effective. B) today's budget concerns make policy makers concerned about whether the money spent on programs is accomplishing the goals intended. C) more psychological programs are being enacted today. D) programs must be tested prior to enactment to ensure the effectiveness of a program. Answer: B Rationale: Today's budget concerns make policy makers concerned about whether the money spent on programs is accomplishing the goals intended, hence driving the need for program evaluation research to assess the effectiveness of these programs. 2) All of the following can be accurately said of program evaluation research EXCEPT which statement? A) Control in program evaluation research is not important. B) Clients who have a vested interested in the program may inflate their ratings of its effectiveness. C) Program evaluation research poses unique practical considerations. D) Program evaluation research often takes place in a complex and uncontrolled natural setting. Answer: A Rationale: Control in program evaluation research is important to ensure valid and reliable results, hence the statement "Control in program evaluation research is not important" is inaccurate. 3) Which of the following is NOT an ethical concern when using program evaluation? A) In programs that meet an urgent need, withholding the program from people in the control group may be denying them effective services. B) People who participate in programs may feel obliged to consent to research in order to continue their benefits. C) Randomly assigning people to a program or a control group ignores the level of need of the person. D) The money spent starting up a program will be wasted if a program evaluation shows the program is ineffective. Answer: D Rationale: While the effectiveness of the program is important, the potential waste of money is not an ethical concern per se. The ethical concerns listed primarily revolve around issues such as fairness, informed consent, and equitable distribution of resources. 4) Which of the following is an accurate statement? A) Program evaluation was popular in the 1960s but is now just about extinct. B) The goal of a good program evaluation is to improve job satisfaction. C) Program evaluation is not a special set of research designs, but rather uses many of the designs presented in the text. D) Program evaluation consists of a special set of research designs which are rarely used for other purposes. Answer: C Rationale: Program evaluation utilizes various research designs presented in research methodologies, and it is not limited to a specific set of designs. Therefore, option C is the accurate statement. 5) One of the most common problems in program evaluation research is: A) finding researchers trained in program evaluation. B) finding laboratory space to conduct the experiments. C) the difficulties encountered in paying participants. D) that such research is often done under stringent ethical and political constraints. Answer: D Rationale: Program evaluation research often involves assessing the effectiveness of interventions or programs implemented in real-world settings, which can be subject to various ethical and political constraints. These constraints can include issues related to privacy, consent, equity, and conflicting stakeholder interests. Negotiating these constraints while maintaining scientific rigor is a common challenge in program evaluation. 6) Program evaluation research is: A) a specific experimental design. B) a group of research methods. C) carried out under high-constraint, laboratory situations. D) diminishing in importance. Answer: B Rationale: Program evaluation research encompasses a variety of research methods used to assess the effectiveness, efficiency, and impacts of programs or interventions. It involves various approaches such as qualitative and quantitative data collection and analysis methods, case studies, surveys, and experimental designs, depending on the nature of the program being evaluated and the research questions being addressed. 7) Program evaluation: A) studies have at least two causal hypotheses. B) involves the use of multiple independent variables. C) modifies existing research designs to meet the particular constraints of the situation. D) should be used only when the employees of the program cannot tell if the program is effective. Answer: C Rationale: Program evaluation often requires adapting or modifying existing research designs to fit the unique context and constraints of the program being evaluated. This flexibility is necessary to ensure that the evaluation methods are appropriate for the specific program and can effectively address the research questions and objectives. 8) A good program evaluator is aware that: A) staff members are interested in doing the best job that they can in applying the program, and evaluating it is often secondary. B) staff members may inadvertently bias data since they wish to present the program in the best possible light. C) often when staff are involved in collecting data, they resent the time that is taken away from other work. D) All of the above Answer: D Rationale: A good program evaluator recognizes that staff members involved in the program may have various motivations and concerns that can impact data collection and evaluation processes. Being aware of potential biases and addressing them through appropriate methods is essential for ensuring the validity and reliability of the evaluation results. 9) More than any other kind of research project, ________ requires a researcher to have excellent political and communication skills. A) laboratory research B) survey research C) educational research D) program evaluation Answer: D Rationale: Program evaluation often involves working with diverse stakeholders, including program administrators, policymakers, funders, and community members. Effective communication and negotiation skills are essential for navigating the political dynamics, managing relationships, and ensuring buy-in and support for the evaluation process and findings. 10) What is the primary problem researchers face in program evaluation? A) Once the evaluation has been completed, recommendations are seldom followed. B) Program evaluation is usually begun before a program starts and before the minor details of the program can be worked out. C) Program evaluation is looked down upon by many administrators and politicians. D) Program evaluation involves attempting high-constraint research in low-constraint naturalistic settings. Answer: D Rationale: Program evaluation often involves conducting research in real-world settings with multiple variables and constraints, which can pose challenges for maintaining methodological rigor and controlling extraneous factors. Balancing the need for rigorous evaluation with the realities of complex, dynamic environments is a primary challenge in program evaluation. 11) More than any other kind of research project, program evaluation requires: A) control of possible confounding variables. B) political and communication skills on the part of the researcher. C) careful analyses of statistical data. D) representativeness of samples. Answer: B Rationale: Program evaluation involves working closely with stakeholders, navigating political dynamics, and communicating findings effectively to various audiences. Therefore, political acumen and strong communication skills are particularly crucial for program evaluators to ensure the success and impact of the evaluation process. 12) Which of the following is a potential biasing factor in gathering data for program evaluation? A) Clients that feel the program has been helpful may inflate effectiveness ratings to insure continued funding. B) Staff members may inflate their ratings of effectiveness because they need to feel they are doing important work and not wasting their time. C) Clients may feel that better programs could be implemented and may therefore deflate their ratings of the current program. D) All of the above Answer: D Rationale: Bias can arise in program evaluation data collection from various sources, including participants' desire to please or receive benefits, staff members' desire to present the program positively, and perceptions of alternative program options. Recognizing and addressing these potential biases are essential for ensuring the integrity and validity of the evaluation findings. 13) In order to check the effects of potential biases in program evaluations, the evaluator should use: A) one well-standardized questionnaire. B) many different data sources, at least some of which are objective. C) observations of neutral observers and not ask the opinions of staff and clients. D) observations of staff only, since they are trained to evaluate the problems of their clients. Answer: B Rationale: Using multiple data sources, including objective measures and different methods of data collection, can help triangulate findings and mitigate the impact of biases in program evaluation. Objective data sources can provide independent verification of subjective information and enhance the credibility and reliability of the evaluation results. 14) Although ________ do NOT indicate the actual effectiveness of a program, they are factors that can influence future effectiveness. A) dependent measures B) unobtrusive measures C) measures of satisfaction D) statistical tests Answer: C Rationale: Measures of satisfaction reflect participants' perceptions and experiences with the program but do not necessarily provide direct evidence of its effectiveness. However, satisfaction levels can impact future engagement, adherence, and outcomes of the program, making them important factors to consider in program planning and improvement efforts. 15) Measurement bias is likely to occur in program evaluation research when A) researchers use objective rather than subjective measures. B) several different dependent measures are used. C) people other than administrators of the program collect the data. D) the data are collected by the same people who run the program. Answer: D Rationale: Measurement bias refers to systematic errors introduced in data collection or analysis. When data are collected by the same individuals who run the program, there's a risk of bias because they may have vested interests in portraying the program in a positive light, leading to potential manipulation or selective reporting of data. 16) Since most programs are developed with several goals in mind, the program evaluator needs to A) rank order the goals and only evaluate the top one or two. B) use several dependent measures to see if the goals are being met. C) use several independent measures to see if the goals are being met. D) employ a particularly sophisticated statistical package in analyzing the data set. Answer: B Rationale: Programs typically have multiple goals, so evaluating only one or two may not provide a comprehensive understanding of program effectiveness. Using several dependent measures allows the evaluator to assess various aspects of program outcomes, providing a more holistic view of its success or shortcomings. 17) Which of the following is NOT an example of a typical dependent measure in program evaluation? A) a questionnaire given to people served by the program, as well as people in the community, asking about their satisfaction with the program B) a measure of politician's commitment to continued funding of the program C) a measure of actual change in the community (e.g., rate of unemployment for a community with a job training program) D) a measure of change in the individual (e.g., a measure of malnutrition for people who receive food stamps) Answer: B Rationale: A measure of politician's commitment to continued funding of the program is not a typical dependent measure in program evaluation because it focuses on external political factors rather than assessing program outcomes or impacts directly. 18) Minimizing bias in dependent measures for program evaluations can be accomplished by using A) objective rather than subjective measures. B) people who are not involved directly in the administration of the program to gather the data. C) several different dependent measures. D) All of the above Answer: D Rationale: Using objective measures reduces the influence of subjective biases. Involving individuals not directly involved in program administration reduces potential conflicts of interest. Additionally, employing multiple dependent measures allows for triangulation of data, enhancing the reliability and validity of the evaluation. 19) Which of the following is NOT a method of control in program evaluations? A) Selecting appropriate dependent measures. B) Minimizing bias in dependent measures. C) Selecting appropriate hypotheses. D) Control through research design. Answer: C Rationale: Selecting appropriate hypotheses is not a method of control in program evaluations; rather, it is a step in formulating the evaluation's research questions and framework. Control in program evaluations is typically achieved through careful selection of dependent measures, minimizing bias, and employing robust research designs. 20) The ideal design (i.e., the design with the most control) for use in program evaluation is A) the randomized control-group design. B) the single-subject, randomized, time-series design. C) the correlational design. D) the reversal or ABA design. Answer: A Rationale: The randomized control-group design provides the highest level of control by randomly assigning participants to experimental and control groups, thus minimizing bias and allowing for causal inferences about program effectiveness. 21) Why would you use a single-group, time-series design instead of a randomized control-group design for a program evaluation? A) A time-series design covers more time and therefore is more accurate. B) A control group may not be ethical or feasible in the program you wish to evaluate. C) The type of data you plan to collect can only be analyzed in terms of a time-series design. D) There is no reason to use the lower-constraint, single-group, time-series design. Answer: B Rationale: Sometimes it may not be ethical or feasible to withhold treatment or intervention from a control group, making a randomized control-group design impractical. In such cases, a single-group, time-series design allows for the examination of changes over time within the same group, providing valuable insights into program effects. 22) The pretest-posttest design is NOT recommended for use in program evaluation research because A) generalization of the results is impossible. B) threats to internal validity are not controlled in this type of design. C) the control group is difficult to debrief. D) it yields low external validity. Answer: B Rationale: The pretest-posttest design is susceptible to threats to internal validity such as history, maturation, and testing effects. Without a control group, it becomes challenging to attribute observed changes solely to the program being evaluated, limiting the validity of the results. 23) In program evaluation, if a control group is NOT possible, the best alternative research strategy is A) a status survey. B) a pretest-posttest design. C) a time-series design. D) a multiple baseline design. Answer: C Rationale: When a control group is not feasible, a time-series design is often the best alternative. This design allows for the comparison of data collected over multiple time points, providing insights into changes within the program without the need for a control group. 24) Which of the following is the weakest design in program evaluation? A) single-group, pretest-posttest B) single-subject C) single-group, time-series D) randomized control group Answer: A Rationale: The single-group, pretest-posttest design is the weakest because it lacks a control group for comparison, making it difficult to attribute observed changes solely to the program being evaluated. This design is susceptible to various threats to internal validity, limiting the strength of causal inferences that can be drawn. 25) Why is the pretest-posttest design NOT recommended for program evaluation? A) With only two measures and no control group, almost none of the threats to reliability are controlled. B) With only two measures, the clients do not have sufficient leeway to express their opinions. C) With only two measures and no control group, almost none of the threats to internal validity are controlled. D) With no control group, there is not enough variability in the experimental group to see an effect. Answer: C Rationale: The pretest-posttest design lacks a control group, which means there is no comparison group to account for factors other than the program itself that may influence the outcomes. Without a control group, threats to internal validity such as history, maturation, and selection cannot be adequately controlled or ruled out, making it difficult to attribute any observed changes solely to the program being evaluated. 26) Which of the following was NOT mentioned in the text as one of the major designs in program evaluation research? A) the status survey B) a randomized control-group design C) a single-group, time-series design D) a pretest-posttest design Answer: A Rationale: The status survey was not mentioned in the text as one of the major designs in program evaluation research. The other options (B, C, D) are all recognized designs commonly used in program evaluation research. 27) The results of the managed care study example in your text A) indicated that the program studied was effective in reducing costs and providing adequate care. B) should be cautiously generalized to other managed care facilities. C) indicated that managed care programs do not effectively treat mental illnesses. D) Both A and B Answer: D Rationale: The results of the managed care study example suggested that the program studied was effective in reducing costs and providing adequate care, but caution should be exercised in generalizing these findings to other managed care facilities. Option D encompasses both of these aspects. 13.4 Surveys 1) Status surveys A) usually focus one or more characteristics of a population. B) are extremely complex. C) cannot be used in natural environments. D) are a very new development. Answer: A Rationale: Status surveys typically focus on describing one or more characteristics of a population. They aim to provide a snapshot of the current status of those characteristics within the population. 2) If a researcher conducts a survey to determine what proportion of college students are registered to vote, the survey is termed a A) status survey. B) political survey. C) fact-finding survey. D) parametric survey. Answer: A Rationale: In this scenario, the survey aims to determine the current status of a specific characteristic (voter registration) within a population (college students), hence it is a status survey. 3) Surveys attempt to obtain information from people A) only in writing. B) in their natural environments. C) primarily in laboratory settings. D) only by telephone. Answer: B Rationale: Surveys aim to gather information from people in their natural environments, which can include various settings such as homes, workplaces, or public spaces. 4) Survey research is a variation of a A) time-series design. B) within-subjects design. C) correlational research design. D) multiple baseline design. Answer: C Rationale: Survey research is a form of correlational research design as it aims to investigate the relationships between variables through the collection of data from a sample of individuals. 5) A survey is a A) comprehensive summary of information gathered. B) proposal for scientific work. C) set of one or more questions presented to participants. D) set of one or more independent measures. Answer: C Rationale: A survey is a method of data collection that involves presenting a set of one or more questions to participants in order to gather information on a particular topic. 6) A survey is most useful in gathering data on A) the knowledge base of participants. B) the attitudes of participants. C) the causes of participants' behavior. D) Both A and B Answer: D Rationale: Surveys can be used to gather data on various aspects including participants' knowledge base and attitudes, making option D the correct choice. 7) The information sought in a status survey is A) a description of the current status of some population characteristic(s). B) a description of the population so a representative sample can be selected. C) the level of socioeconomic status. D) a ranking of importance of a certain set of items. Answer: A Rationale: A status survey aims to describe the current status of one or more characteristics within a population, as indicated in option A. 8) Which of the following is an example of a status survey? A) a survey to determine whether teachers' job satisfaction is related to teacher training B) a survey to determine what proportion of voters are Republican, Democrat, Liberal, or Independent C) a survey to determine whether socioeconomic status is related to education level D) a survey to determine the correlation between hours of sleep and performance on exams. Answer: B Rationale: A status survey aims to describe characteristics of a population at a specific point in time. Option B, a survey to determine the proportion of voters belonging to different political affiliations, fits this description by seeking to ascertain the current status of a population characteristic (political affiliation). 9) Which of the following statements is NOT accurate? A) Survey research is relatively new compared to status surveys. B) Survey research is more complex than status surveys. C) Survey research tries to discover relationships among variables. D) Survey research does not seek the current status of population characteristics. Answer: D Rationale: Survey research is designed to gather information about various aspects of a population, including its current characteristics, attitudes, opinions, behaviors, and relationships among variables. Thus, option D is not accurate as survey research does seek the current status of population characteristics. 10) Survey research is most similar to which type of research design? A) experimental B) correlational C) differential D) case-study Answer: B Rationale: Survey research, like correlational research, aims to identify relationships among variables without manipulation or control over them. It involves collecting data from a sample to understand the associations between different factors. 11) The major goal of a survey is to A) learn about how people behave generally. B) learn about how people will actually behave in certain circumstances. C) learn about the ideas, knowledge, feelings, opinions, attitudes, and self-reported behavior of a defined population of people by directly asking them. D) None of the above Answer: C Rationale: Surveys are specifically designed to gather information directly from individuals within a defined population by asking them questions about their thoughts, opinions, attitudes, behaviors, and other relevant aspects. 12) Which of the following is NOT a major step in survey research? A) Define the population to be studied. B) Decide how the survey is to be administered. C) Determine what area of information is to be sought. D) Randomly assign participants to groups. Answer: D Rationale: Randomly assigning participants to groups is a step commonly associated with experimental research designs, not survey research. In survey research, participants are typically sampled from the population of interest rather than randomly assigned to groups. 13) Which of the following is NOT a major step in survey research? A) Develop a sampling frame and obtain a representative sample. B) Pretest the first draft of the survey with a subsample and refine it. C) Use control procedures to ensure experimenter objectivity. D) Analyze, interpret, and communicate the results. Answer: C Rationale: Control procedures to ensure experimenter objectivity are more relevant to experimental research designs where manipulation and control over variables are central. In survey research, the focus is on gathering data from participants through structured instruments such as questionnaires or interviews. 14) What is an interview schedule? A) a calendar of times available for interviewing B) a method of randomly assigning participants to interview times C) a specific set of questions that is asked of participants D) interviews that are conducted at certain intervals Answer: C Rationale: An interview schedule refers to a predetermined set of questions that an interviewer asks respondents during an interview. It helps maintain consistency and structure in data collection. 15) Which of the following are two types of survey instruments? A) questionnaires and interview schedules B) status and descriptive C) survey research and survey experiments D) None of the above Answer: A Rationale: Questionnaires and interview schedules are two common types of survey instruments used to collect data from respondents. Questionnaires are typically self-administered, while interview schedules involve direct interaction between the interviewer and the respondent. 16) What kind of survey question seeks information about the respondent, such as age, sex, occupation, and family status? A) status B) demographic C) content D) graphic Answer: B Rationale: Demographic survey questions seek information about the characteristics of the respondents, such as age, gender, occupation, education level, and family status. These questions help researchers characterize the sample and understand how different demographic factors may relate to other variables of interest. 17) Content items on questionnaires ask about A) the respondents' ability to understand the content of the questionnaire. B) sex, age, occupation, and family status of the respondent. C) respondents' opinions, attitudes, knowledge, and behavior. D) demographic characteristics of the respondent. Answer: C Rationale: Content items on questionnaires refer to the substantive questions that inquire about respondents' opinions, attitudes, knowledge, behavior, or other relevant aspects of the research topic. They aim to gather information beyond mere demographic characteristics. 18) In a survey questionnaire, a question that asks "What is the capital of North Dakota?" is a A) demographic question. B) content item. C) geological item. D) capitalistic question. Answer: B Rationale: A content item refers to a question that directly addresses the subject matter or content being surveyed. In this case, the question about the capital of North Dakota directly pertains to geographical knowledge, making it a content item rather than a demographic, geological, or economic inquiry. 19) Which of the following is NOT a step in developing a survey instrument? A) Determine what content area the questions should cover. B) Determine the form and order of the questions to be asked. C) Determine the form of reimbursement for the respondent. D) Determine the procedure that will be used to administer the survey (e.g., telephone survey, face-to-face interview). Answer: C Rationale: Determining the form of reimbursement for respondents is not typically a step in developing a survey instrument. While it may be relevant in certain cases, it's not a fundamental step in the design and construction of the survey itself, unlike determining the content, form and order of questions, and the administration procedure. 20) Which of the following is NOT a characteristic of a good questionnaire? A) It begins with an introduction, which explains the purpose of the survey and gives instructions. B) The language is clear and concise and is appropriate for the population being studied. C) It has an uncluttered and orderly appearance and is well within the reading abilities of the respondents. D) In group administration settings, the order of the questions are changed to prevent respondents from reading or copying the answers of the person next to them. Answer: D Rationale: Changing the order of questions in group administration settings is not a characteristic of a good questionnaire. It can introduce bias or disrupt the intended flow of the survey. A good questionnaire ensures clarity, readability, and coherence of questions, but deliberately altering question order to prevent copying is not a recommended practice. 21) How do surveys impose some constraints on participants? A) by making the participants "blind" to the purpose of the questions B) by setting time limits for the questionnaire C) by using a specific survey instrument D) by giving an impression of the answers sought to the respondent Answer: C Rationale: Surveys impose constraints on participants by using a specific survey instrument, which dictates the format, types of questions, and response options available to participants. This constraint ensures consistency and comparability in responses across participants. 22) A typical survey includes A) demographic questions. B) content items. C) descriptive questions. D) Both A and B. Answer: D Rationale: A typical survey often includes both demographic questions (pertaining to characteristics of respondents) and content items (questions directly related to the subject matter being surveyed). Therefore, option D, "Both A and B," is correct. 23) The most information and generally the best results are obtained when a survey is administered A) by telephone. B) by mail. C) incognito. D) in a personal, face-to-face interview. Answer: D Rationale: Personal, face-to-face interviews typically yield the most information and best results in survey administration. This method allows for rapport-building, clarification of questions, and observation of non-verbal cues, all of which can enhance data quality. 24) Survey research should A) be used in early exploratory research. B) use a "shotgun" approach to cover as many areas as possible. C) have a clear focus and be guided by hypotheses and expectancies held by the researcher. D) All of the above Answer: C Rationale: Survey research should have a clear focus and be guided by hypotheses and expectancies held by the researcher to ensure relevance, coherence, and effectiveness. Using a "shotgun" approach (option B) can lead to unfocused data collection, while survey research is indeed suitable for early exploratory research (option A), but it should still maintain a clear focus. 25) If you ask participants in a survey a question such as "What is your opinion of the way the economy is being handled by the president?", this is an example of A) an open-ended question. B) a multiple-choice question. C) a Likert scale question. D) None of the above Answer: A Rationale: This question prompts participants to provide their opinion in their own words, without being constrained by predefined response options. Therefore, it qualifies as an open-ended question, where respondents have the freedom to express their thoughts and sentiments. 26) Before forming the questions of a survey, the researcher should identify A) the general area of inquiry. B) the population to be surveyed. C) the form of administration. D) All of the above Answer: D Rationale: Before forming survey questions, researchers should identify the general area of inquiry, the population to be surveyed, and the form of administration to ensure that the questions are relevant, appropriate, and effectively administered. 27) Which of the following is NOT a type of item found on a survey? A) multiple-choice B) open-ended C) Myasthenia scale D) Likert scale Answer: C Rationale: The Myasthenia scale is not a typical type of item found on a survey. Instead, surveys commonly include multiple-choice questions, open-ended questions, and Likert scale questions to gather a variety of types of responses from participants. 28) In Likert scales, A) respondents rate the degree to which they felt they correctly answered the previous item. B) the items are arranged on a continuum, with extreme positions at the end points. C) the respondent indicates the degree to which they agree with or "like" the content of the scale. D) the researcher measures the degree to which the items on a survey are correlated. Answer: B Rationale: Likert scales present respondents with a series of statements or items that they are asked to rate based on the degree of agreement or disagreement. These items are typically arranged on a continuum, with options ranging from strongly disagree to strongly agree, allowing respondents to indicate their level of agreement or disagreement. 29) When the ________ about which a researcher seeks information is (are) large and diverse, a ________ should be used a represent it. A) population; sample B) sample; population C) participants; respondent D) respondents; participant Answer: A Rationale: When the population about which a researcher seeks information is large and diverse, a sample should be used to represent it. Sampling involves selecting a subset of individuals from the larger population to study, allowing researchers to make inferences about the population based on the characteristics of the sample. 30) The population in a survey is A) the group of respondents that were given the survey but did not complete it. B) the country about which we wish to obtain information. C) the larger group about whom we wish to obtain information. D) the group of respondents who were given the survey and completed it. Answer: C Rationale: The population in a survey refers to the larger group about whom researchers wish to obtain information. It represents the entire group of interest, from which a sample is drawn to conduct the survey. 31) A researcher wishes to survey the opinions of science teachers in the United States about science policy. He/she gives the questionnaires to all the biology teachers in the Minneapolis high schools. What is the population in this study? A) Science teachers in the United States. B) Biology teachers in the United States. C) Teachers in Minneapolis. D) High school teachers in Minneapolis. Answer: A Rationale: The population in this study is science teachers in the United States because the researcher aims to survey the opinions of science teachers across the entire country, not just in Minneapolis. 32) A researcher wishes to survey the opinions of science teachers in the United States about science policy. He/she gives the questionnaire to all the biology teachers in the Denver high schools. What is the sample in this study? A) High school teachers in Denver. B) Biology teachers in Denver high schools. C) Science teachers in the United States. D) Science teachers in Denver high schools. Answer: B Rationale: The sample in this study is biology teachers in Denver high schools because they represent a subset of the larger population of science teachers in the United States. 33) The U.S. census is an example of A) a survey of a representative sample. B) a survey that attempts to question every member of a population. C) an interview survey. D) None of the above Answer: B Rationale: The U.S. census is an example of a survey that attempts to question every member of a population, making it a comprehensive enumeration rather than a sampling survey. 34) The sample in survey research must be drawn in such a way as to A) adequately represent the population to which we want to generalize. B) be able to randomly assign participants to groups. C) minimize researcher bias. D) minimize the variance. Answer: A Rationale: The sample in survey research must be drawn in such a way as to adequately represent the population to which we want to generalize. This ensures that findings from the sample can be generalized to the larger population with confidence. 35) If a researcher wishes to have confidence in inferences from a sample to a population, the sample must A) represent adequately the population to which he/she wants to generalize. B) be small enough not to incorporate the entire population. C) be large enough to incorporate the entire population. D) represent populations other than the one about which he/she wishes to generalize. Answer: A Rationale: If a researcher wishes to have confidence in inferences from a sample to a population, the sample must represent adequately the population to which he/she wants to generalize. This ensures that findings from the sample accurately reflect the characteristics of the population. 36) The heart of survey research is A) statistical analysis. B) the careful selection of representative samples. C) a commitment to the democratic system. D) getting a quick public response to an issue. Answer: B Rationale: The heart of survey research lies in the careful selection of representative samples. This ensures that the data collected accurately reflect the characteristics and opinions of the population of interest, allowing researchers to make valid inferences. 37) Carrying out a survey by interviewing the first 100 people coming out of a movie theater would be an example of A) a status survey. B) nonprobability sampling. C) probability sampling. D) simple random sampling. Answer: B Rationale: Nonprobability sampling involves techniques where the probability of each member of the population being selected for the sample is not known. Interviewing the first 100 people coming out of a movie theater does not ensure equal probability of selection for all members of the population, as it depends on who happens to exit the theater first. Thus, it falls under nonprobability sampling. 38) The chief advantage of nonprobability sampling is that A) the results are not biased. B) it constitutes a representative sample. C) it is easy to carry out. D) it is expensive compared to other methods. Answer: C Rationale: Nonprobability sampling methods are relatively easier to carry out compared to probability sampling methods. They often involve convenience or judgmental sampling, which requires less planning and resources. 39) Random sampling is a type of ________ sampling. A) nonprobability B) probability C) nonrepresentative D) none of the above Answer: B Rationale: Random sampling, whether simple random sampling, systematic random sampling, or stratified random sampling, is a type of probability sampling where each member of the population has a known probability of being selected for the sample. 40) Interviewing the first 50 people who walk into a store is an example of A) probability sampling. B) simple random sampling. C) stratified random sampling. D) nonprobability sampling. Answer: D Rationale: Interviewing the first 50 people who walk into a store is an example of nonprobability sampling because it does not ensure equal probability of selection for all members of the population. The sample depends on who happens to enter the store first, which may not be representative of the entire population. 41) What is the major disadvantage of nonprobability sampling? A) It is difficult to carry out. B) The sample may not be representative of the total population. C) The statistical procedures needed are not easily available. D) The robustness of the sample is decreased. Answer: B Rationale: The major disadvantage of nonprobability sampling is that the sample may not be representative of the total population. This is because the selection of individuals for the sample is often based on convenience or judgment, rather than ensuring every member of the population has an equal chance of being included. 42) In probability sampling, A) a representative sample of the population is not obtained. B) each member of the population is included in the sample. C) each member of the population has some known probability of being included in the sample. D) Both A and C Answer: C Rationale: In probability sampling, each member of the population has some known probability of being included in the sample, ensuring that the sample is representative of the population. 43) In simple random sampling, every member of the identified population A) has an equal chance of being selected. B) has an unequal probability of being selected. C) has a simple relationship to the population. D) is willing to participate in the research. Answer: A Rationale: Simple random sampling ensures that every member of the identified population has an equal chance of being selected for the sample, without any bias. 44) The initial list (such as all children in a school district) from which we draw a sample is termed A) the sample size. B) sampling error. C) a stratum. D) a sample frame. Answer: D Rationale: The initial list from which a sample is drawn is called the sample frame. It includes all the members of the population from which the sample will be selected. 45) What is a sampling frame? A) a method used to sample a population B) a method used to stratify a population for sampling C) a list of all the members in a population D) the series of decisions necessary to create a sample Answer: C Rationale: A sampling frame is a list of all the members in a population from which a sample will be drawn. It serves as the basis for selecting the sample. 46) Stratified random sampling procedures are used when we wish to A) represent accurately subgroups of a population in the sample. B) study geological events. C) employ a larger sample. D) take a random sample from the entire population. Answer: A Rationale: Stratified random sampling is used when we wish to represent accurately subgroups of a population in the sample. It ensures that each subgroup is represented proportionally in the sample, allowing for more precise analysis of each subgroup. 47) What is a major problem with simple random sampling? A) The researcher needs a list of all the members of a population, which is difficult to obtain for a large population. B) The researcher needs to contact all the people in the population, which would be difficult for a large population. C) The sample usually over-represents certain segments of the population. D) There is a systematic bias that leads to persons with certain characteristics having a higher probability of being selected. Answer: A Rationale: Simple random sampling requires a complete list of all members of the population, which can be challenging to obtain, especially for large populations. Without such a list, it becomes difficult to ensure every member has an equal chance of being selected. 48) Stratified random sampling procedures are used when A) certain subgroups are more prevalent in some samples than others. B) it is important to ensure that subgroups within a population are adequately represented in the sample. C) the population being sampled is homogeneous. D) it is important to ensure that certain subgroups of a population are not included in the sample. Answer: B Rationale: Stratified random sampling is utilized when it's crucial to ensure that various subgroups within the population are adequately represented in the sample. This method helps to maintain proportionality and accuracy in representing diverse segments of the population. 49) If you want to get a representative sample for a survey, you should use a A) probability sample. B) nonprobability sample. C) nonprobability sample from stratified groups. D) None of the above. Answer: A Rationale: Probability sampling methods ensure that every member of the population has a known and non-zero chance of being selected, thus providing a representative sample for the survey. 50) The advantage of a stratified random sample over a random sample is that A) stratified random samples require less work than random samples. B) stratified random samples are more likely to represent accurately a population, because stratification on key variables virtually assures accurate representation on those variables. C) the random sample is not a probability sample. D) stratified random samples provide both a probability and a nonprobability sample in one procedure. Answer: B Rationale: Stratified random sampling enhances the accuracy of representation by ensuring that key variables are proportionately represented within the sample, thus reducing the risk of bias and improving the generalizability of findings to the population. 51) The size of the sample needed depends on the ________ of the population. A) attitudes B) homogeneity C) mean D) demand characteristics Answer: B Rationale: The homogeneity of the population influences the required sample size. More homogeneous populations typically require smaller sample sizes compared to heterogeneous populations to achieve sufficient representation. 52) In general, if the population is ________, then ________ sample sizes are needed. A) small; larger B) heterogeneous; smaller C) homogeneous; smaller D) homogeneous; larger Answer: C Rationale: Homogeneous populations require smaller sample sizes because there is less variability within the population, reducing the need for a large sample to capture diverse characteristics adequately. 53) In order to determine the confidence limits for estimating population characteristics, we need to estimate A) the number of participants we will need. B) how much time the survey will take. C) all of the population parameters. D) the size of the standard deviation in the population for the characteristics we wish to measure. Answer: D Rationale: To establish confidence limits, we need to estimate the standard deviation of the population characteristics we aim to measure. This information is crucial for determining the precision of our estimates and constructing confidence intervals. 54) A researcher states she is 95 percent confident that the population mean is between 50 and 75. The confidence interval is A) 95 percent. B) 5 percent. C) 50-75. D) 60 plus or minus 5. Answer: C Rationale: The confidence interval represents the range within which we are confident the population parameter (in this case, the population mean) lies. In this scenario, the researcher is 95% confident that the population mean falls between 50 and 75. 55) What is the most important characteristic of any sample? A) It should be homogeneous. B) It should accurately represent the population. C) It should be heterogeneous. D) It should include the entire population. Answer: B Rationale: The primary goal of sampling is to obtain a sample that accurately represents the population of interest. Therefore, the most important characteristic of any sample is its ability to provide an accurate reflection of the population. 56) In general, ________ samples represent populations better than do ________ samples. A) heterogeneous; homogeneous B) homogeneous; heterogeneous C) smaller; larger D) larger; smaller Answer: D Rationale: Larger samples tend to represent populations better than smaller samples because they provide more data points, reducing the margin of error and improving the accuracy of estimates. 57) A homogeneous population is A) one in which the members are similar to one another. B) one in which there is more diversity. C) a population of people genetically related to homosexuals. D) a population of genetically related people. Answer: A Rationale: A homogeneous population consists of members who share similar characteristics or traits. This similarity makes it easier to generalize findings to the entire population because there is less variability among individuals. 58) In which type of population is it more important to have a large sample? A) a homogeneous population B) a heterogeneous population C) a stratified population D) a random population Answer: B Rationale: In a heterogeneous population, there is more variability among individuals. Therefore, a larger sample size is needed to adequately capture this diversity and ensure that the sample is representative of the population as a whole. 59) The smaller the sample size in a survey study, the A) greater our confidence in the results. B) larger the confidence intervals around the population estimates. C) smaller the confidence intervals around the population estimates. D) more difficult it is to develop the survey instrument. Answer: B Rationale: With a smaller sample size, there is greater uncertainty in the estimates obtained from the sample. This leads to larger confidence intervals around the population estimates, indicating less precision in the findings. 60) How do researchers determine what size sample is necessary? A) By estimating the variability of the characteristic of the population that they want to measure. B) By estimating the mean of the characteristic of the population that they want to measure. C) By estimating the reliability of the survey instrument. D) Both A and B Answer: A Rationale: Researchers determine the necessary sample size by considering the variability of the characteristic they want to measure within the population. Higher variability requires a larger sample size to ensure that the sample accurately represents the population. 61) A cross-sectional survey design involves administering the survey A) only once to a particular sample. B) to the same group of people more than once. C) to people whose names are on file as volunteer respondents. D) None of the above Answer: A Rationale: Cross-sectional surveys are conducted once, capturing data from a specific sample at a single point in time to provide a snapshot of the population's characteristics. 62) Longitudinal survey designs are also referred to as A) cross-sectional designs. B) panel designs. C) between-subjects designs. D) threshold designs. Answer: B Rationale: Longitudinal survey designs involve collecting data from the same group of participants over multiple time points, allowing researchers to observe changes and trends over time. These are commonly referred to as panel designs. 63) Which of the following is NOT a basic design used in survey research? A) cross-sectional design B) longitudinal design C) experimental design D) panel design Answer: C Rationale: Experimental design is not typically considered a basic design in survey research. It involves manipulating variables to observe their effects, which is distinct from purely observational survey methods. 64) In a cross-sectional design, the survey is administered A) twice to each participant, yielding a difference score that indicates the amount of change over time. B) once to each participant, yielding data on the measured characteristics as they exist at the time of the survey. C) many times to each participant, with an intervening manipulation that is predicted to have an effect on the survey. D) None of the above Answer: B Rationale: In a cross-sectional design, the survey is administered once to each participant to gather data on the measured characteristics as they exist at the time of the survey, providing a snapshot of the population. 65) In a longitudinal design, the survey is administered A) to the same group of participants successively at different times. B) many times to each participant, with an intervening manipulation that is predicted to have an effect on the survey. C) once to each participant, yielding data on the measured characteristics as they exist at the time of the survey. D) to different groups of people at different times to allow comparisons to be made of population characteristics. Answer: A Rationale: In a longitudinal design, the same group of participants is surveyed at multiple time points to track changes and developments over time, allowing researchers to observe trends and patterns within the same group. 66) Provide Rationales for each answer. Maintain the same format of the question followed by the answer Followed by the Rationale. Mention "Rationale:" Above the rationale. Include original questions as well as the other options. Answer: D Rationale: Longitudinal designs are ideal for studying changes over time within the same group of individuals. By administering surveys successively at different times, researchers can observe how variables change or remain stable within the group, providing valuable insights into developmental processes, trends, and causal relationships over time. 67) What is a common problem with longitudinal designs? A) Attrition (participants dropping out) B) Participants may be unwilling to complete the survey several times. C) Long-term studies are often expensive. D) All of the above Answer: D Rationale: Attrition, participant unwillingness to complete multiple surveys, and the high cost associated with long-term studies are all common challenges encountered in longitudinal designs. These factors can affect the validity and reliability of the study findings. 68) A longitudinal design is essentially a A) between-subjects design. B) within-subjects design. C) between-groups design. D) within-groups design. Answer: B Rationale: A longitudinal design involves studying the same subjects repeatedly over a period of time, making it essentially a within-subjects design as it focuses on changes within individuals over time. 69) A variation that falls between the cross-sectional and longitudinal design is the A) single-measures survey design. B) sequential survey design. C) semi-group survey design. D) single-group random-assignment survey design. Answer: B Rationale: The sequential survey design involves studying different cohorts over time, allowing for comparisons at different points in time, which bridges the gap between cross-sectional and longitudinal designs. 70) The sequential survey design allows A) comparisons at different points in time. B) highly biased sampling. C) the avoidance of the need to sample. D) easy interpretations of causality. Answer: A Rationale: The sequential survey design facilitates comparisons at different points in time, enabling researchers to observe changes or trends over time within different cohorts. 71) The sequential survey design is essentially A) a within-subjects design. B) a study without participants. C) naturalistic observation. D) a cross-sectional survey repeated over time. Answer: D Rationale: The sequential survey design involves conducting cross-sectional surveys at different points in time with different cohorts, allowing for comparisons over time. 72) An appropriate design for a business to use to determine the best marketing strategies given changes in the market is a A) within-subjects design. B) sequential survey design. C) single-subject design. D) a single-group, pretest-posttest design Answer: B Rationale: A sequential survey design would be appropriate for studying changes in market trends over time by examining different cohorts and their responses to various marketing strategies. 13.5 Ethical Principles 1) People may feel obliged to participate in research when they are receiving benefits from a program that is being evaluated. This subtle form of coercion violates a participant's right to A) privacy. B) informed consent. C) free services. D) Both A and B Answer: D Rationale: Both privacy and informed consent are violated when participants feel obliged to participate due to receiving benefits from a program under evaluation. This situation undermines their autonomy and ability to make a voluntary decision to participate. 2) What is one of the reasons why program evaluation research poses significant ethical problems? A) The programs were created by politicians. B) The program staff may not approve of the evaluation. C) It would be impossible to identify a control group. D) The research participants are typically recipients of the services of the program and therefore may feel obliged to participate. Answer: D Rationale: Program evaluation research often involves participants who are also recipients of program services, creating a potential conflict of interest where participants may feel compelled to participate, compromising their autonomy and voluntary consent. 3) Why is informed consent such a tricky issue in program evaluation? A) Participants often feel a subtle coercion to consent to an evaluation because they are receiving services through the program. B) Most of the people served by such programs would not be legally able to give consent. C) In general, the program evaluator could not give the participants sufficient information for informed consent without jeopardizing the validity of the evaluation. D) All of the above Answer: A Rationale: Informed consent is challenging in program evaluation because participants may feel coerced due to receiving program services, which compromises their ability to make a voluntary and informed decision to participate. 4) When might a randomized control-group design be unethical in a program evaluation? A) whenever there are more people needing service than the program has funds to serve B) whenever it would mean denying services for which there is a critical need. C) whenever the evaluation will take more than a few weeks to complete D) whenever we are dealing with children Answer: B Rationale: A randomized control-group design might be unethical in program evaluation when it involves denying critical services to participants, especially when there are more people in need than the program can serve. This raises ethical concerns about withholding necessary interventions for the sake of research. 5) What kind of privacy concerns might raise ethical issues in a program evaluation? A) Programs are funded by government agencies and therefore do not provide confidentiality to their clients. B) The researcher must obtain information from the general public as a control group, and these people have not consented to receive services from the program. C) The participants will be asked to reveal information to the evaluator that would normally be confidential. D) The information gathered in a program evaluation is required to be made available to anyone who requests it through the Freedom of Information Act. Answer: C Rationale: In option C, the concern is about breaching confidentiality by asking participants to disclose sensitive information. This raises ethical issues because it violates the trust between the evaluator and the participants, potentially leading to harm or discomfort for the individuals involved. Protecting the privacy of participants is a fundamental ethical principle in research and program evaluation. 6) One of the ethical difficulties in evaluating mental health programs is A) that the subjects are emotionally disturbed. B) randomizing to treatment and control groups. C) a typical lack of participants. D) a typical lack of funding. Answer: B Rationale: Option B is correct because randomizing subjects to treatment and control groups in mental health programs can pose ethical challenges. Randomization may result in some individuals not receiving potentially beneficial treatment, which raises concerns about fairness and the well-being of participants. Balancing the ethical imperative to provide effective treatment with the methodological necessity of randomization can be a complex issue in mental health program evaluation. 7) In evaluating a nutritional breakfast program for children in poverty, randomly assigning children to breakfast/no-breakfast groups A) will create clear statistical problems. B) will create data that are uninterpretable. C) is an ethical problem. D) should always include a repeated measures component. Answer: C Rationale: Option C is correct because randomly assigning children to breakfast/no-breakfast groups in a nutritional breakfast program for children in poverty raises ethical concerns. Denying breakfast to some children, particularly those in need, could lead to negative consequences for their health and well-being. This violates the principle of beneficence and may cause harm to vulnerable populations. 8) Polls conducted by political organizations _________ provide information about the poll's validity. A) usually provide perfect B) always provide C) should provide D) should not provide Answer: C Rationale: Option C is correct because political organizations conducting polls should provide information about the poll's validity as part of ethical research practices. Transparency about methodology, sampling techniques, and potential biases is essential for evaluating the credibility and reliability of poll results. Providing such information allows the public to assess the validity of the findings and make informed judgments. Test Bank for Research Methods: A Process of Inquiry Anthony M. Graziano, Michael L. Raulin 9780205900923, 9780205907694, 9780135705056