CH-App-D Selecting Statistical Procedures – Research Methods: A Process of Inquiry : Exam Guide

Chapter 18 Appendix D: Selecting Statistical Procedures 18.1 Appendix D Items 1) The decision about how to analyze the data A) should be made before the data are collected. B) should be made after the data are collected. C) should be made randomly. D) is not a particularly important concern in research. Answer: A Rationale: Option A is correct because determining the data analysis approach before data collection ensures that appropriate data collection methods are employed, thereby aligning with the research objectives and facilitating meaningful analysis. 2) The type of model recommended in the text to help the researcher in determining appropriate statistical tests is a A) decision-tree model. B) mathematical model. C) statistical model. D) spread sheet model. Answer: A Rationale: Option A is correct because a decision-tree model provides a systematic framework for researchers to navigate through various factors such as research hypotheses, variables, and data characteristics to select appropriate statistical tests. 3) The authors of the text suggest a ________ to help the beginning researcher determine appropriate statistical procedures. A) branch-tree model B) decision-tree model C) statistical model D) statistical text Answer: B Rationale: Option B is correct because a decision-tree model helps beginning researchers by presenting a structured flowchart or decision-making process to guide them in selecting suitable statistical procedures based on specific characteristics of their research design. 4) The t-test for independent groups is typically used with A) score data from two independent samples. B) any kind of categorical data. C) score data from non-independent samples. D) any kind of numerical data. Answer: A Rationale: Option A is correct because the t-test for independent groups compares the means of two independent samples, making it suitable for analyzing numerical score data from two distinct groups or conditions. 5) The decision regarding how to analyze data should be made A) in the idea-generation phase. B) after the data has been collected. C) before the data has been collected. D) when the study has been accepted for publication. Answer: C Rationale: Option C is correct because deciding how to analyze data before data collection ensures that appropriate data collection methods are chosen to yield meaningful results aligned with the research objectives. 6) Which is NOT a characteristic used to determine appropriate statistical procedures? A) number of independent variables B) research hypotheses C) age of participants D) level of measurement for each dependent variable Answer: C Rationale: Option C is correct because while the age of participants may influence study design, it is not directly related to selecting appropriate statistical procedures, unlike other factors such as variables, research hypotheses, and levels of measurement. 7) What is the dependent variable and the level of measurement of the dependent variable in a study in which you compare participants (matched for age) on how induced anxiety affects the speed at which they recognize words? A) speed of word recognition; ratio B) age; interval C) speed of word recognition; interval D) age; ratio Answer: A Rationale: Option A is correct because the dependent variable is the speed of word recognition, which is measured in time (e.g., seconds), making it a ratio-level variable. 8) In more complex research, in which there may be several research hypotheses and several dependent measures, A) an analysis of variance (ANOVA) will typically suffice. B) a t-test is sufficient, no matter what the complexities are. C) selection of appropriate statistical tests is better deferred until the data are collected. D) we may need a number of statistical procedures to analyze the data. Answer: D Rationale: Option D is correct because complex research designs with multiple hypotheses and dependent measures may require a combination of statistical procedures, such as ANOVA, regression analysis, and multivariate techniques, to adequately analyze the data and address research questions. 9) Which of the following characteristic(s) of a study is(are) needed to determine the appropriate statistical test? A) nature of dependent variable(s) B) research hypotheses C) number and levels of independent variables D) All of the above Answer: D Rationale: Option D is correct because all the listed characteristics (nature of dependent variables, research hypotheses, number and levels of independent variables) are essential considerations for determining the appropriate statistical test in a research study. 10) In a study using the gender of the participant as an independent variable, how many levels of the independent variable are there? A) 1 B) 2 C) 3 D) None, since sex cannot be manipulated Answer: B Rationale: Option B is correct because gender typically has two levels (male and female) and is considered a categorical variable with two distinct categories. 11) A study hypothesizes gender differences in mathematical ability. Boys and girls in the fifth and the sixth grade are given a math test that includes addition, subtraction, and division. What is the dependent variable? A) gender of participant B) score on math test C) type of question D) grade in school Answer: B Rationale: Option B is correct because the dependent variable in this study is the score achieved by participants on the math test. It is what the study seeks to measure and assess regarding potential gender differences in mathematical ability. 12) Boys and girls in the fifth and the sixth grade are given a math test that includes addition, subtraction, and division. What is a possible research hypothesis. A) Boys and girls differ in mathematical ability. B) Fifth and sixth graders differ in mathematical ability. C) There is a gender X type of question interaction; Boys are better in addition and subtraction and girls are better in division. D) All of the above Answer: D Rationale: Option D is correct because all the provided statements represent possible research hypotheses that could be tested based on the described study scenario. 13) The first step in the decision model suggested in the text is to A) determine the type of data. B) determine the number of research hypotheses. C) determine the level of statistical significance. D) describe the research so the major characteristics of the study can be identified. Answer: D Rationale: Option D is correct because the first step in the decision model involves describing the research to identify its major characteristics, which helps in subsequent decision-making regarding appropriate statistical procedures. 14) The decision-tree model presented in the book is based on three basic steps necessary for determining appropriate statistical procedures. Which of the following is NOT one of the basic steps? A) Describe the research. B) Identify important characteristics. C) Decide level of significance. D) Based on identified characteristics, make decisions to arrive at statistical test. Answer: C Rationale: Option C is correct because deciding the level of significance is not one of the basic steps in the decision-tree model for determining appropriate statistical procedures. The level of significance is typically set based on research conventions and hypotheses. 15) STUDY: A researcher is interested in genetic links to panic disorder. She selects a sample of 500 people from a Psychology 101 class. Each participant is given a structured interview to determine if they have panic disorder. Two groups are created (panic disorder and no panic disorder). All participants are given a questionnaire to determine the number of relatives who have anxiety disorders. The researcher calculates if the group of participants with panic disorder differs from the group of participants without panic disorder in the number of relatives with anxiety disorders. What is the independent variable in the above study? A) panic disorder diagnosis B) number of relatives with anxiety disorders C) score on the questionnaire D) None of the above Answer: A Rationale: Option A is correct because the independent variable in the study is the presence or absence of panic disorder diagnosis among the participants. 16) STUDY: A researcher is interested in genetic links to panic disorder. She selects a sample of 500 people from a Psychology 101 class. Each participant is given a structured interview to determine if they have panic disorder. Two groups are created (panic disorder and no panic disorder). All participants are given a questionnaire to determine the number of relatives who have anxiety disorders. The researcher calculates if the group of participants with panic disorder differs from the group of participants without panic disorder in the number of relatives with anxiety disorders. What is the level of measurement of the dependent measure in the above study? A) nominal B) ordinal C) interval D) ratio Answer: D Rationale: Option D is correct because the number of relatives with anxiety disorders is measured on a quantitative scale, making it a ratio-level variable. 17) STUDY: A researcher is interested in genetic links to panic disorder. She selects a sample of 500 people from a Psychology 101 class. Each participant is given a structured interview to determine if they have panic disorder. Two groups are created (panic disorder and no panic disorder). All participants are given a questionnaire to determine the number of relatives who have anxiety disorders. The researcher calculates if the group of participants with panic disorder differs from the group of participants without panic disorder in the number of relatives with anxiety disorders. What type of data is generated for the dependent measure(s) in the above study? A) nominal B) score C) ordered D) All of the above Answer: B Rationale: Option B is correct because the dependent measure (number of relatives with anxiety disorders) yields numerical scores, making it a type of score data. 18) STUDY: A researcher is interested in genetic links to panic disorder. She selects a sample of 500 people from a Psychology 101 class. Each participant is given a structured interview to determine if they have panic disorder. Two groups are created (panic disorder and no panic disorder). All participants are given a questionnaire to determine the number of relatives who have anxiety disorders. The researcher calculates if the group of participants with panic disorder differs from the group of participants without panic disorder in the number of relatives with anxiety disorders. What is (are) the research hypothesis (or hypotheses) in the study described above? A) There are two research hypotheses: that there are people with panic disorder and that they have relatives with anxiety disorders. B) There is one research hypothesis: that people have panic disorder. C) There is one research hypothesis: that people with panic disorder differ in the number of relatives with anxiety disorders from people who do not have panic disorder. D) There is one research hypothesis: that people who have relatives with anxiety disorders will have panic disorder. Answer: C Rationale: Option C is correct because the research hypothesis states a specific expectation about the relationship between panic disorder and the number of relatives with anxiety disorders, indicating a difference between participants with and without panic disorder. 19) STUDY: A researcher is interested in genetic links to panic disorder. She selects a sample of 500 people from a Psychology 101 class. Each participant is given a structured interview to determine if they have panic disorder. Two groups are created (panic disorder and no panic disorder). All participants are given a questionnaire to determine the number of relatives who have anxiety disorders. The researcher calculates if the group of participants with panic disorder differs from the group of participants without panic disorder in the number of relatives with anxiety disorders. Which of the following statistical procedures is appropriate for the study above? A) chi square B) t-test C) correlation coefficient D) MANOVA Answer: B Rationale: Option B is correct because a t-test is appropriate for comparing the means of two independent groups, which is the case in this study where the researcher wants to determine if the number of relatives with anxiety disorders differs between participants with and without panic disorder. 20) STUDY: A major airline wants to know if pilots make significantly more errors and more serious errors when they are under a particular stress (e.g., if their airplane is hijacked while in flight). The airline also wants to know if pilots' years of experience is a factor. The following research is set up. A 767 flight simulator is used. All of the pilot's behavior is monitored, so that the number of errors made by the pilot can be easily read off the computer record. The seriousness of each error is judged later by master pilots using a 25-point scale (interval) of seriousness. The participants are jetliner pilots with experience ranging from 1 to 20 years of commercial piloting. The participants are randomly assigned to two conditions, and the years of experience is the same for both groups. All pilots simulate take-off from New York, a two-and-a-half hour flight at 32,000 feet cruising altitude, and a landing at Miami. The pilots are told that their job rating depends on how well they handle this simulated flight, and it is impressed on them that they are to do everything as if this were a real flight. In the experimental condition, two masked "passengers" break into the cockpit 45 minutes into the flight and, aiming guns at the pilots, demand to be taken to Miami for refueling, a payment of a million dollars, and then, flown to Cuba. They threaten to kill some of the passengers and then the pilots and themselves if anything goes wrong. The control group has no "hijacking" and proceeds to Miami without incident. How many levels of the manipulated independent variable are there in the study described above? A) 1 B) 2 C) 3 D) There is no manipulated independent variable. Answer: B Rationale: Option B is correct because the study involves two conditions: experimental (hijacking scenario) and control (no hijacking scenario), making it a two-level manipulated independent variable design. 21) STUDY: A major airline wants to know if pilots make significantly more errors and more serious errors when they are under a particular stress (e.g., if their airplane is hijacked while in flight). The airline also wants to know if pilots' years of experience is a factor. The following research is set up. A 767 flight simulator is used. All of the pilot's behavior is monitored, so that the number of errors made by the pilot can be easily read off the computer record. The seriousness of each error is judged later by master pilots using a 25-point scale (interval) of seriousness. The participants are jetliner pilots with experience ranging from 1 to 20 years of commercial piloting. The participants are randomly assigned to two conditions, and the years of experience is the same for both groups. All pilots simulate take-off from New York, a two-and-a-half hour flight at 32,000 feet cruising altitude, and a landing at Miami. The pilots are told that their job rating depends on how well they handle this simulated flight, and it is impressed on them that they are to do everything as if this were a real flight. In the experimental condition, two masked "passengers" break into the cockpit 45 minutes into the flight and, aiming guns at the pilots, demand to be taken to Miami for refueling, a payment of a million dollars, and then, flown to Cuba. They threaten to kill some of the passengers and then the pilots and themselves if anything goes wrong. The control group has no "hijacking" and proceeds to Miami without incident. What is the level of constraint in the study described above? A) naturalistic B) experimental C) correlational D) differential Answer: B Rationale: The study described involves manipulation of an independent variable (presence or absence of hijacking scenario) to observe its effects on dependent variables (number and seriousness of errors). This experimental setup allows for causal inferences to be drawn, indicating a higher level of constraint compared to correlational or naturalistic designs. 22) STUDY: A researcher is interested in genetic links to panic disorder. She selects a sample of 500 people from a Psychology 101 class. Each participant is given a structured interview to determine if they have panic disorder. Two groups are created (panic disorder and no panic disorder). All participants are given a questionnaire to determine the number of relatives who have anxiety disorders. The researcher calculates if the group of participants with panic disorder differs from the group of participants without panic disorder in the number of relatives with anxiety disorders. How many dependent variables are there in the above study? A) 4 B) 3 C) 2 D) 1 Answer: D Rationale: In the study described, the dependent variable is the number and seriousness of errors made by the pilots. There is only one dependent variable being measured - the frequency and severity of errors. 23) Descriptive statistics may be sufficient A) in higher-constraint research. B) in lower-constraint research. C) to rule out confounding variables. D) to determine answers to causal questions. Answer: B Rationale: Descriptive statistics may suffice in lower-constraint research where the focus is primarily on describing the data rather than making causal inferences. In such cases, researchers may aim to characterize patterns or relationships without manipulating variables. 24) STUDY: A major airline wants to know if pilots make significantly more errors and more serious errors when they are under a particular stress (e.g., if their airplane is hijacked while in flight). The airline also wants to know if pilots' years of experience is a factor. The following research is set up. A 767 flight simulator is used. All of the pilot's behavior is monitored, so that the number of errors made by the pilot can be easily read off the computer record. The seriousness of each error is judged later by master pilots using a 25-point scale (interval) of seriousness. The participants are jetliner pilots with experience ranging from 1 to 20 years of commercial piloting. The participants are randomly assigned to two conditions, and the years of experience is the same for both groups. All pilots simulate take-off from New York, a two-and-a-half hour flight at 32,000 feet cruising altitude, and a landing at Miami. The pilots are told that their job rating depends on how well they handle this simulated flight, and it is impressed on them that they are to do everything as if this were a real flight. In the experimental condition, two masked "passengers" break into the cockpit 45 minutes into the flight and, aiming guns at the pilots, demand to be taken to Miami for refueling, a payment of a million dollars, and then, flown to Cuba. They threaten to kill some of the passengers and then the pilots and themselves if anything goes wrong. The control group has no "hijacking" and proceeds to Miami without incident. What are the dependent measures in the study described above? A) Years of experience and number of errors. B) Flight time to Miami and number of errors. C) Flight time, number and seriousness of errors. D) Number and seriousness of errors. Answer: D Rationale: The dependent measures in the study are the number and seriousness of errors made by the pilots. These variables are being observed and measured to assess the impact of the experimental manipulation (presence or absence of hijacking scenario) on pilot performance. 25) STUDY: A major airline wants to know if pilots make significantly more errors and more serious errors when they are under a particular stress (e.g., if their airplane is hijacked while in flight). The airline also wants to know if pilots' years of experience is a factor. The following research is set up. A 767 flight simulator is used. All of the pilot's behavior is monitored, so that the number of errors made by the pilot can be easily read off the computer record. The seriousness of each error is judged later by master pilots using a 25-point scale (interval) of seriousness. The participants are jetliner pilots with experience ranging from 1 to 20 years of commercial piloting. The participants are randomly assigned to two conditions, and the years of experience is the same for both groups. All pilots simulate take-off from New York, a two-and-a-half hour flight at 32,000 feet cruising altitude, and a landing at Miami. The pilots are told that their job rating depends on how well they handle this simulated flight, and it is impressed on them that they are to do everything as if this were a real flight. In the experimental condition, two masked "passengers" break into the cockpit 45 minutes into the flight and, aiming guns at the pilots, demand to be taken to Miami for refueling, a payment of a million dollars, and then, flown to Cuba. They threaten to kill some of the passengers and then the pilots and themselves if anything goes wrong. The control group has no "hijacking" and proceeds to Miami without incident. What is the independent variable in the above study? A) years of experience B) hijacking C) number of errors D) flight time to Miami Answer: B Rationale: The independent variable in the study is the presence or absence of a hijacking scenario. This is the factor that is manipulated by the researchers to observe its effects on the dependent variables (number and seriousness of errors). 26) Which of the following is NOT an important consideration in using decision-tree flowchart? A) We must describe the research. B) We must identify the dependent variables. C) We must present our findings in an appropriate journal. D) We must identify the independent variables along with their number of levels. Answer: C Rationale: While presenting findings in a journal is important, it is not directly related to the process of constructing a decision-tree flowchart. The other options are crucial steps in utilizing a decision-tree flowchart for organizing research methodology and data analysis. 27) Which of the following is NOT important to specify when using the decision-tree model? A) levels of the independent variable B) the dependent variables C) the research hypotheses D) debriefing of research participants Answer: D Rationale: Debrieifng of research participants is not typically part of specifying the decision-tree model. It's important to specify the independent variable levels, dependent variables, and research hypotheses to ensure clarity and accuracy in the research design and analysis. 28) STUDY: A major airline wants to know if pilots make significantly more errors and more serious errors when they are under a particular stress (e.g., if their airplane is hijacked while in flight). The airline also wants to know if pilots' years of experience is a factor. The following research is set up. A 767 flight simulator is used. All of the pilot's behavior is monitored, so that the number of errors made by the pilot can be easily read off the computer record. The seriousness of each error is judged later by master pilots using a 25-point scale (interval) of seriousness. The participants are jetliner pilots with experience ranging from 1 to 20 years of commercial piloting. The participants are randomly assigned to two conditions, and the years of experience is the same for both groups. All pilots simulate take-off from New York, a two-and-a-half hour flight at 32,000 feet cruising altitude, and a landing at Miami. The pilots are told that their job rating depends on how well they handle this simulated flight, and it is impressed on them that they are to do everything as if this were a real flight. In the experimental condition, two masked "passengers" break into the cockpit 45 minutes into the flight and, aiming guns at the pilots, demand to be taken to Miami for refueling, a payment of a million dollars, and then, flown to Cuba. They threaten to kill some of the passengers and then the pilots and themselves if anything goes wrong. The control group has no "hijacking" and proceeds to Miami without incident. What is the level of measurement of the dependent measures in the above study? A) ratio; interval B) nominal; ratio; interval C) interval; ordinal D) interval; ordinal; ratio Answer: A Rationale: The level of measurement for the dependent measures (number and seriousness of errors) is ratio. This level of measurement allows for meaningful interpretation of ratios, differences, and proportions among the measured values. 29) STUDY: A major airline wants to know if pilots make significantly more errors and more serious errors when they are under a particular stress (e.g., if their airplane is hijacked while in flight). The airline also wants to know if pilots' years of experience is a factor. The following research is set up. A 767 flight simulator is used. All of the pilot's behavior is monitored, so that the number of errors made by the pilot can be easily read off the computer record. The seriousness of each error is judged later by master pilots using a 25-point scale (interval) of seriousness. The participants are jetliner pilots with experience ranging from 1 to 20 years of commercial piloting. The participants are randomly assigned to two conditions, and the years of experience is the same for both groups. All pilots simulate take-off from New York, a two-and-a-half hour flight at 32,000 feet cruising altitude, and a landing at Miami. The pilots are told that their job rating depends on how well they handle this simulated flight, and it is impressed on them that they are to do everything as if this were a real flight. In the experimental condition, two masked "passengers" break into the cockpit 45 minutes into the flight and, aiming guns at the pilots, demand to be taken to Miami for refueling, a payment of a million dollars, and then, flown to Cuba. They threaten to kill some of the passengers and then the pilots and themselves if anything goes wrong. The control group has no "hijacking" and proceeds to Miami without incident. What type(s) of data is(are) generated for the dependent measure(s) in the above study? A) nominal; score; ordered B) score; interval C) nominal; score D) score; score Answer: D Rationale: The dependent measures (number and seriousness of errors) generate score data, which are numerical values representing the observed variables. In this study, there are no nominal or ordered categories associated with the dependent measures. 30) STUDY: A researcher is interested in genetic links to panic disorder. She selects a sample of 500 people from a Psychology 101 class. Each participant is given a structured interview to determine if they have panic disorder. Two groups are created (panic disorder and no panic disorder). All participants are given a questionnaire to determine the number of relatives who have anxiety disorders. The researcher calculates if the group of participants with panic disorder differs from the group of participants without panic disorder in the number of relatives with anxiety disorders. What type of design is the research described above? A) correlational B) experimental C) differential D) naturalistic Answer: C Rationale: The described research involves comparing two groups (experimental condition with hijacking scenario and control group without hijacking scenario) to determine if there is a difference in the dependent variable (number and seriousness of errors). This is indicative of a differential design where groups are compared based on the presence or absence of an experimental manipulation. 31) Which of the following is NOT a measure of central tendency? A) mean B) median C) variance D) mode Answer: C Rationale: Variance is not a measure of central tendency; instead, it is a measure of dispersion or spread in a dataset. 32) Which statement is the most accurate? A) For each research hypothesis, there is only one appropriate statistical test. B) For each research hypothesis, there may be more than one appropriate statistical test, although they generally lead to the same conclusions. C) For each research hypothesis, there are many appropriate statistical tests that always lead to the same conclusions. D) For each research hypothesis, there are two appropriate statistical tests, and they generally lead to different conclusions. Answer: B Rationale: Different research hypotheses may require different statistical tests based on factors such as the type of data, research design, and assumptions. While there might be multiple appropriate tests, they generally lead to similar conclusions if applied correctly. 33) To use the flowchart in Appendix D, we need to A) first complete the study. B) make sure the data analyses are complete. C) provide informed consent feedback to all participants. D) describe the study. Answer: D Rationale: The flowchart in Appendix D likely outlines steps related to designing or planning a study. To utilize it effectively, one needs to describe the study first to determine the appropriate steps to follow. 34) Before carrying out ________ statistics, it is helpful to calculate ________ statistics. A) parametric; nonparametric B) descriptive; preliminary C) inferential; descriptive D) descriptive; inferential Answer: C Rationale: Before performing inferential statistics (which draw conclusions and make inferences about populations based on sample data), it is helpful to calculate descriptive statistics (which summarize and describe characteristics of the sample data). 35) If a researcher wished to determine if two groups (e.g., children and adults) differed statistically on a measure, he/she would use A) inferential statistics. B) primary statistics. C) descriptive statistics. D) secondary statistics. Answer: A Rationale: Inferential statistics are used to draw conclusions about populations based on sample data, which would be appropriate for comparing groups statistically. 36) If a researcher wished to present the demographic characteristic of a sample, he/she would use A) inferential statistics. B) primary statistics. C) descriptive statistics. D) secondary statistics. Answer: C Rationale: Descriptive statistics are used to summarize and present the characteristics of a sample, including demographic information. 37) Which of the following is NOT listed in the text as an example of secondary analyses? A) data snooping B) post hoc analyses C) ex post facto analyses D) analyses designed to help explain the pattern of results Answer: C Rationale: Secondary analyses typically involve utilizing existing data for purposes other than those for which the data were originally collected. Ex post facto analyses, on the other hand, involve examining data collected after an event has occurred for explanatory purposes and are not typically considered secondary analyses. 38) A researcher examines four groups of participants on the effects of diet (two conditions: high protein and normal) and exercise (two levels: modest and moderate) on grades. Participants are exposed to one of four combinations of the independent variables and are measured once on the dependent variable, which produces score data. The proper statistical procedure would be a(n) A) one-way ANOVA. B) factorial ANOVA. C) repeated measures ANOVA. D) correlated t-test. Answer: B Rationale: A factorial ANOVA would be appropriate for analyzing the effects of two independent variables (diet and exercise) with multiple levels on a continuous dependent variable (grades). 39) The statistical test appropriate to test for mean differences among more than two groups is A) a t-test for independent groups. B) a chi-square test. C) a correlated t-test. D) an analysis of variance (ANOVA). Answer: D Rationale: ANOVA (Analysis of Variance) is used to test for mean differences among more than two groups. 40) STUDY: A major airline wants to know if pilots make significantly more errors and more serious errors when they are under a particular stress (e.g., if their airplane is hijacked while in flight). The airline also wants to know if pilots' years of experience is a factor. The following research is set up. A 767 flight simulator is used. All of the pilot's behavior is monitored, so that the number of errors made by the pilot can be easily read off the computer record. The seriousness of each error is judged later by master pilots using a 25-point scale (interval) of seriousness. The participants are jetliner pilots with experience ranging from 1 to 20 years of commercial piloting. The participants are randomly assigned to two conditions, and the years of experience is the same for both groups. All pilots simulate take-off from New York, a two-and-a-half hour flight at 32,000 feet cruising altitude, and a landing at Miami. The pilots are told that their job rating depends on how well they handle this simulated flight, and it is impressed on them that they are to do everything as if this were a real flight. In the experimental condition, two masked "passengers" break into the cockpit 45 minutes into the flight and, aiming guns at the pilots, demand to be taken to Miami for refueling, a payment of a million dollars, and then, flown to Cuba. They threaten to kill some of the passengers and then the pilots and themselves if anything goes wrong. The control group has no "hijacking" and proceeds to Miami without incident. Given the four hypotheses below, what statistical procedures are appropriate to use? Hypothesis 1: There will be more errors in the hijack condition. Hypothesis 2: Seriousness will be greater in the hijack condition. Hypothesis 3: In the hijack condition there will be an inverse relationship between years of experience and number of errors. Hypothesis 4: In the hijack condition, there will be an inverse relationship between years of experience and seriousness of errors. A) ANOVA for all of the hypotheses B) ANOVA; ANOVA; correlation coefficient; correlation coefficient C) t-test; Pearson r; Spearman r; chi-square D) chi-square; Spearman r; Pearson r; t-test Answer: B Rationale: • Hypotheses 1 and 2 involve comparing means between two conditions (hijack and control), making ANOVA appropriate. • Hypotheses 3 and 4 involve examining the relationship between years of experience and errors/seriousness, which can be analyzed using correlation coefficients. 41) Post hoc analyses A) are unplanned and therefore should not be used. B) can be useful in elaborating the meaning of planned analyses. C) are completed after the research article has been published. D) are completed after secondary analyses have been completed. Answer: B Rationale: Post hoc analyses are conducted after the primary planned analyses to explore unexpected results or to delve deeper into the findings. They are not inherently unplanned; rather, they are conducted as an additional step to enhance understanding and interpretation of the data. 42) What are two major measures of variability of data? A) variance and standard deviation B) mean and median C) mode and alpha D) None of the above Answer: A Rationale: Variance and standard deviation are both measures of variability, providing insights into how spread out the data points are from the mean. Mean and median are measures of central tendency, not variability. 43) When research produces unexpected results, ________ analyses may be used to help explain the results. A) discrepancy B) ancillary C) filibuster D) secondary Answer: D Rationale: Secondary analyses are often utilized to explore unexpected results further, helping researchers to understand the underlying reasons behind these outcomes. 44) Research articles should present descriptive statistics on the demographic characteristics of their sample to A) allow comparisons with other studies in terms of the populations being studied. B) give information needed to determine the limits of generalizability for the study. C) present information on the possible biases of the sample (e.g., all college students). D) All of the above. Answer: D Rationale: Providing demographic characteristics allows for a comprehensive understanding of the sample, aiding in comparisons with other studies, assessing generalizability, and identifying potential biases. 45) Which two statistical tests can give us the same information? A) t-test and two-way ANOVA B) Spearman r and chi-square C) inferential and descriptive statistics D) t-test and one-way ANOVA Answer: D Rationale: Both t-test and one-way ANOVA can be used to compare means between groups, hence providing similar information in certain contexts. 46) A researcher does a study on the effects of rap music on reaction time. The participants are matched on initial reaction time, then one of each pair is assigned randomly to either the experimental or control group, and his/her matched partner is automatically assigned to the other group. To analyze this study, ________ should be used. A) a repeated measures ANOVA B) a t-test for independent groups C) a correlated t-test D) either A or C Answer: D Rationale: Since participants are matched and assigned to different groups, either a repeated measures ANOVA or a correlated t-test would be appropriate for analyzing the data. 47) STUDY: A researcher is interested in genetic links to panic disorder. She selects a sample of 500 people from a Psychology 101 class. Each participant is given a structured interview to determine if they have panic disorder. Two groups are created (panic disorder and no panic disorder). All participants are given a questionnaire to determine the number of relatives who have anxiety disorders. The researcher calculates if the group of participants with panic disorder differs from the group of participants without panic disorder in the number of relatives with anxiety disorders. What is the dependent measure in the above study? A) number of relatives with anxiety disorders B) panic disorder diagnosis C) result of the structured interview D) gender Answer: A Rationale: The dependent measure in this study is the number of relatives with anxiety disorders, as it is what the researcher is comparing between the two groups (panic disorder and no panic disorder). 48) STUDY: A major airline wants to know if pilots make significantly more errors and more serious errors when they are under a particular stress (e.g., if their airplane is hijacked while in flight). The airline also wants to know if pilots' years of experience is a factor. The following research is set up. A 767 flight simulator is used. All of the pilot's behavior is monitored, so that the number of errors made by the pilot can be easily read off the computer record. The seriousness of each error is judged later by master pilots using a 25-point scale (interval) of seriousness. The participants are jetliner pilots with experience ranging from 1 to 20 years of commercial piloting. The participants are randomly assigned to two conditions, and the years of experience is the same for both groups. All pilots simulate take-off from New York, a two-and-a-half hour flight at 32,000 feet cruising altitude, and a landing at Miami. The pilots are told that their job rating depends on how well they handle this simulated flight, and it is impressed on them that they are to do everything as if this were a real flight. In the experimental condition, two masked "passengers" break into the cockpit 45 minutes into the flight and, aiming guns at the pilots, demand to be taken to Miami for refueling, a payment of a million dollars, and then, flown to Cuba. They threaten to kill some of the passengers and then the pilots and themselves if anything goes wrong. The control group has no "hijacking" and proceeds to Miami without incident. What type of design is represented in the research described above? A) correlated-groups design B) independent-groups design C) nonparametric design D) Both B and C Answer: B Rationale: The research described represents an independent-groups design, where participants are randomly assigned to either the experimental (hijacking) or control (no hijacking) condition. 49) Data snooping is A) referring to data or research that has not been published. B) changing the hypothesis of a study after the data have been collected and analyzed. C) used most often in large scale, lower-constraint research when clues to potential relationships between variables may be buried in the mass of data. D) used to control confounding variables in experimental research. Answer: C Rationale: Data snooping refers to the practice of exploring data extensively to find significant results, often leading to spurious findings due to multiple comparisons or fishing expeditions, which can compromise the integrity of the research. 50) Which of the following is NOT a type of secondary analysis? A) ex post facto analyses B) post hoc analyses C) explanatory analyses D) data snooping Answer: A Rationale: Ex post facto analyses involve studying data after events have occurred or conditions have changed, typically without the researcher's control. This is not considered a type of secondary analysis, as secondary analysis generally involves re-analyzing existing data for different research purposes. 51) What is the primary value of post hoc tests? A) They tell us that at least one mean is significantly different from at least one other mean. B) They tell us which means are significantly different from which other means. C) They reduce the total number of statistical tests necessary. D) They help us in selecting the appropriate statistical analysis. Answer: B Rationale: Post hoc tests are conducted after an analysis of variance (ANOVA) to determine which specific group means are significantly different from each other. They provide valuable information about pairwise comparisons among groups, helping to identify where significant differences lie. 52) It is often the case that secondary analyses may far outnumber the primary analyses in A) experimental research studies. B) higher-constraint research studies. C) lower-constraint research studies. D) research on vision. Answer: C Rationale: Lower-constraint research studies, which typically have fewer pre-specified hypotheses and more exploratory analyses, often result in a higher number of secondary analyses compared to primary analyses. 53) What type of secondary analyses should be included in the report of any research project, regardless of its level of constraint? A) analyses associated with "data snooping" B) only secondary analyses that help explain the results C) all post hoc analyses D) descriptive statistics on the demographic characteristics of the sample Answer: D Rationale: Regardless of the level of constraint in a research study, it is essential to include descriptive statistics on the demographic characteristics of the sample to provide a comprehensive overview of the study participants. 54) Data snooping is most useful in A) large-scale, lower constraint studies. B) small-scale, experimental studies. C) all higher constraint research. D) tightly controlled and meticulously executed research designs. Answer: A Rationale: Data snooping, which involves exploring data for unexpected patterns or relationships, is most useful in large-scale, lower constraint studies where there may be a plethora of variables and relationships to explore. 55) Monte Carlo studies continue to show the remarkable robustness of most statistics to assumption violations, particularly A) when the sample sizes are large. B) when the sample sizes are equal. C) for nonparametric statistics. D) when the sample sizes are small. Answer: B Rationale: Monte Carlo studies simulate various sampling scenarios and analyze the behavior of statistics under different conditions. They often reveal that statistics are remarkably robust to assumption violations, especially when sample sizes are equal. 56) If an assumption of a statistical test can be violated without threatening the validity of the conclusions drawn, the test is said to be ________ to violations of that assumption. A) withstanding B) robust C) strong D) powerful Answer: B Rationale: A statistical test that can withstand violations of its assumptions without compromising the validity of its conclusions is termed as robust. 57) In a study in which scores on a math test are being used as a dependent variable, the researcher discovers that the scores are positively skewed. This violates the assumption of normality for the t-test that the researcher planned to use. The best solution for the researcher is to A) not use the t-test because the results may not be valid. B) use another test that does not have the same assumptions. C) check if the t-test is robust to violations of normality. D) prorate the data so it fits a normal curve. Answer: C Rationale: Before discarding the use of the t-test, the researcher should first investigate whether the test is robust to violations of normality, as many parametric tests are robust to moderate violations of assumptions. 58) If a data set violates an assumption of the statistical test that is to be used, the test should be ________ to this violation if we want to be confident that our conclusions are valid. A) robust B) withstanding C) strong D) powerful Answer: A Rationale: For confidence in the validity of conclusions drawn from a statistical test, it's important that the test is robust to violations of its assumptions, meaning it can tolerate such violations without significantly affecting the results. 59) The series of studies that showed many statistical tests to be robust to violations of assumptions on which they are based are known as A) Reno studies. B) robust studies. C) violation studies. D) Monte Carlo studies. Answer: D Rationale: Monte Carlo studies are specifically designed to test the robustness of statistical methods under various conditions, including violations of assumptions. 60) Which of the following is an important consideration in the robustness of many statistical tests? A) equal sample sizes B) normality of the sample C) randomness of the sample D) demographics of the sample Answer: A Rationale: Equal sample sizes contribute to the robustness of statistical tests, as they help stabilize the estimation process and reduce the impact of outliers or skewed distributions. 61) The conclusion that most parametric statistics are robust to violations of most of their assumptions is based on a series of computer simulations studies known as A) Monte Carlo studies. B) Las Vegas studies. C) the Jacquard method. D) the Mann-Whitney studies. Answer: A Rationale: Monte Carlo studies involve generating random samples from known distributions and applying statistical methods to assess their performance under various conditions. These simulations allow researchers to systematically test the robustness of statistical methods to violations of assumptions. 62) In which category is ANOVA properly classified? A) nonparametric statistics B) parametric statistics C) descriptive statistics D) inferential and nonparametric statistics Answer: B Rationale: ANOVA (Analysis of Variance) is a parametric statistical method used to analyze differences between means of multiple groups. It assumes that the data are normally distributed and that the groups have equal variances. 63) When an underlying assumption of an inferential statistical test can be violated without threatening the validity of the conclusion, the statistical test is said to be A) hardy in all conditions. B) robust to violations of the assumption. C) inappropriate for use. D) meaningless. Answer: B Rationale: A statistical test is considered robust if it can produce valid results even when its underlying assumptions are not perfectly met. This means that minor violations of assumptions do not significantly affect the accuracy or reliability of the test. 64) Most parametric statistics are ________ violations of assumptions about population distributions. A) sensitive to B) ignorant of C) robust to D) redolent of Answer: C Rationale: Parametric statistics are often robust to violations of assumptions about population distributions, meaning that they can still provide valid results even when the data do not perfectly adhere to the assumptions of the statistical test. This robustness makes parametric methods widely applicable in practice. Test Bank for Research Methods: A Process of Inquiry Anthony M. Graziano, Michael L. Raulin 9780205900923, 9780205907694, 9780135705056