Research Methods in Psychology International Edition 9th Edition by David G – Test Bank

 

Instant Download – Complete Test Bank With Answers

 

 

Sample Questions Are Posted Below

 

CHAPTER 6

RELATIONAL RESEARCH

 

Synopsis

This chapter examines research that assesses how variables are related to each other. The chapter first considers contingency research, and then focuses on correlational research (techniques that use coefficients of correlation for analysis). These modes of observations are discussed in detail here, because they receive little attention elsewhere in the book, insofar as they are used to gather data. Correlation coefficients are discussed as statistical tools. The chapter ends with a discussion of how experimentation allows statements about causation–at least in principle.

 

Outline

CONTINGENCY RESEARCH

CORRELATIONAL RESEARCH

The Correlation Coefficient

Interpreting Correlation Coefficients

Low Correlations: A Caution

Reactivity in Correlational Research

Regression

Complex Correlational Procedures

Experimentation and Internal Validity

Application: Causation

SUMMARY

KEY CONCEPTS

EXERCISES

SUGGESTED RESOURCE

WEB RESOURCES

LAB RESOURCES

PSYCHOLOGY IN ACTION:  Amount of Sleep and Tension Headaches

 

New to This Edition

• Updated Organization: This chapter was Chapter 5 in the previous edition.

 

• Updated Examples: Contingency “table” research (as it was called in the previous edition) has been changed to “contingency research” when that type of research is referred to. It has also been added that this type of research is usually at the nominal level of measurement. In some contingency tables, incorrect numbers have been corrected.

 

The discussion of regression has been expanded with an additional example. This new example concerns the predictors associated with using weapons in cases of domestic violence.

 

In the Web Resources section at the end of the chapter, Langston has been updated to Langston (2011) and the associated chapter numbers have been changed.

• Updated & Additional References: The reference to Rushton (1997) has been deleted. Plous (1998) has been added to expand the discussion comparing animal-rights activists and non-activists in the section on contingency tables. Rushton, Ankney, & Davidson (2009) has been added to update the section on correlation between brain size and cognitive ability. Kernsmith & Craun (2008) has been added to the Regression section as an additional example that concerns the predictors associated with using weapons in cases of domestic violence. Elmes & Lorig (2008) has been added to a paragraph at the end of the chapter that discusses internal validity and experimental control.

 

Definitions of Key Concepts

A cell is an entry within a particular combination of a row and a column in a data (e.g., contingency) table.

 

Contingency research examines how the frequency of observations changes as a function of two nominal variables.

 

Correlational research is a type of relational research that permits a determination of both the degree and direction of a relationship between variables.

 

A correlation coefficient is a statistic (that can vary from -1 to 1, and) that reflects the nature of the relationship between two variables. The coefficient’s sign indicates the direction of the relationship between variables, and the magnitude of the difference from zero indicates the degree, or strength, of that relationship.

 

Covariates are variables that also change along with a target variable in a study. These factors often are accounted for within statistical analyses, such as regression analyses in correlational research.

 

A cross-lagged panel correlation procedure measures the correlation between variables at several different times. The procedure allows time for particular relationships to develop, and can enhance internal validity.

 

Ex post facto means after the fact. In relational research data are related in this fashion because the variables are not directly manipulated.

 

Multiple regression refers to a regression procedure that predicts changes along a target variable as a function of several other, correlated variables (also see regression).

 

Pearson’s product-moment correlation coefficient (Pearson r) represents one type of correlation coefficient for evaluations of the relationship between two variables.

 

Proximate causes are the immediate causes of an event that derive from manipulations in an experiment.

 

 

A regression procedure uses a correlation coefficient to predict changes along one variable as a function of changes along one or more other variables.

 

Relational research attempts to determine how two or more variables are related in the absence of a direct manipulation of those variables.

 

Scatter diagrams are graphical depictions of the degree of a correlation between variables. Scores along each variable are displayed along different axes, and each data point reflects the combination of individual data along both variables.

 

A truncated range refers to a lack of variability in the data along a variable, which can result in a low correlation coefficient and generally complicate the interpretation of low coefficients.

 

Ultimate causes are the final (theoretical) explanations of an event that prevent additional questions concerning why the given event occurred.

 

The c² test for independence is a statistical test that determines whether there is a difference between predicted (e.g., by chance) and obtained data. This test often is used to determine whether or not there is a statistically significant relationship between variables specified by data in a contingency table.

 

Answers to Exercises

1. [Special Exercise]. The negligible correlation probably resulted from a truncated range on the intelligence-test scores. Since the participants were all college graduates, it is likely that the range in scores was not very great. Only a heterogeneous sample of intelligence-test scores will yield a meaningful r.

 

2. Included in Table 6-3 (next page) are the numbers needed to calculate Pearson r for the data listed in column (b) and column (c) of Table 6-3. The values for X (head size in cm) are identical for both examples, and are displayed in the left panel of Table 6-3. The differing values for Y are separately displayed for column (b) (middle panel) and (c) (right). Summed values are included along the bottom rows of the table. To obtain r, these sums can be plugged directly in the following formula from Box A-1 of Appendix A: r = nSXY – (SX)( SY) / Ö{[nSX² – (SX)²][ nSY² – (SY)²]}.

 

Table 6-3. Values for calculations of r from Table 6-3 of the text.

 

3. The calculated correlation between verbal SAT scores and freshman GPA was quite high and positive (r = .688, p < .05). Thus, there appears to be a strong relationship between the two scores; as verbal SAT scores increase, so do freshman GPA’s. Researchers would probably conclude that the verbal SAT score is a fairly strong predictor of how well a person will do in their coursework freshman year.

 

4. [Special Exercise]. The text identifies several problems with relational research along with potential solutions. The most primary of these problems is the threat to internal validity; it is difficult to establish causation in relational work because variables are not manipulated. Obtaining a strong correlation does not indicate a causal relationship between variables since confounding variables could be responsible for the observed relation. Internal validity could potentially be improved by the use of a cross-lagged panel correlation procedure, where the repeated determination of a correlation might reduce the likelihood of confounding. The amount of reactivity is typically not known in relational research either, though reactivity could be minimized by not collecting measures along both variables at the same time. In this way, the participant will not be as likely to recognize that the researcher is interested in a relationship between the variables of interest. The reliance on a truncated range of data along a given variable complicates the interpretation of low correlation coefficients. Using heterogeneous populations can reduce the likelihood of a truncated range. Failing to meet the assumptions of the statistical test (e.g., the data reflect a curvilinear, rather than a linear, relationship between variables) also can complicate the interpretation of coefficients. Researchers can construct a scatter plot to see if linearity is violated and, if so, can conduct trend analyses instead of a simple correlation.

 

Suggestions for Discussion

Use of c². You might devote some class time to the c² statistic, especially as it is used as a test for independence. In statistics classes students usually have trouble with the null hypothesis for the statistic, usually because both the independence and goodness-of-fit versions are considered at the same time. This should not be a problem here if discussion is limited to the independence use of the statistic.

Plous’ work on animals-rights activists has plenty of data that can be analyzed with the independence test.

Application: Causation. Students generally understand that causation cannot be easily established in correlational research because the variables are not manipulated. However, students often do not realize that this problem arises out of confounding from other variables. In order to highlight this point, students can be asked to produce other factors that could be responsible for a previously obtained correlation (e.g., between smoking and lung cancer). By subsequently developing ways to address these potential confounds (perhaps by experiment rather than relational study), the concept of proximate cause can be introduced as a specific assertion from research (e.g., No statistically significant relationship was found between food preferences and incidence of cancer). Finally, this concept can be differentiated from ultimate cause by contrasting specific statements of proximate cause with broad assertions on the topic that are not bound to a particular study or method (e.g., Likelihood of cancer is genetically pre-determined).

 

Psychology in Action: Amount of Sleep and Tension Headaches. The research on tension headaches and amount of sleep described in the Psychology in Action section is a good project to have students undertake in courses with a laboratory component, because it combines survey with relational methods. You can obtain a substantial amount of data very quickly. If you have your class do this project, have them obtain some additional demographic data (such as, age, year in school, sex, major, and so on). The additional data could be used for more detailed correlational analyses, but they also can be used as quasi-independent variables for a project described at the end of chapter 10. Since the original research was analyzed as a ² design, you might want to indicate how that statistic is calculated. Treating the data as interval/ratio allows some repetition of the measurement scale concept discussed in chapter 3.

Alternatively, in lecture or discussion courses students could collectively be asked to develop both a contingency table and a correlational design to evaluate the nature of a relationship between two variables (e.g., memory span and test performance). Instructors could use the generated designs to reveal that contingency tables rely on frequencies of categorized behaviors that may be determined by the researcher, whereas correlations tend to rely on data from interval or ratio scales.

 

Langston (2011), Research Methods Laboratory Manual for Psychology. Chapters 4 and 5 of the manual are concerned with correlation research. Included is a brief description of the meaning of a correlation coefficient for students seeking additional reinforcement of this concept. Also provided are target articles that evaluate a probability misjudgment hypothesis and the “gap problem.” Suggested projects for a laboratory component include replication studies and investigative research for additional factors and variables.

 

Experimental Dilemma

A researcher in a small southern town was interested in the relation between the citizen’s religious beliefs and their attitudes towards pornography. The researcher enlisted the help of the two town ministers (a Baptist and a Methodist), who were concerned about the amount of pornography present in movies and magazines.

Members of the two congregations were administered a long survey that assessed their religiosity and their attitudes towards pornography. Previous work with this survey had shown that it is a reliable and valid indicator of religious attitudes, and it is a valid assessor of beliefs about pornography. A total of 430 adults were surveyed, which represented about 75% of the town’s adult population. The researcher assessed the correlation between religiosity and attitudes towards pornography and found r = -.03. The researcher concluded that this small, essentially zero, correlation between the two variables indicates that they are not correlated–one’s views about pornography are unrelated to one’s religious beliefs. Do you agree or disagree with the researcher’s conclusion?  Why or why not?

Answer. The student should disagree. The small r probably arose from a truncated range on both variables. That is, most of these people were likely to be fairly high on religiosity, and they were likely to have uniform attitudes towards pornography (it makes no difference whether these attitudes were pro or con with regard to the outcome of the correlation). The researcher should have tried to obtain a random sample of people in the community, so that wider ranges of attitudes on the two dimensions could have been assessed. Another potential difficulty with this project is reactivity. Even if the respondents were guaranteed that their responses would be anonymous, they would all be aware of the purpose of the research and the fact that their ministers were interested in the outcome. Therefore, the responses may have been socially desirable ones (at least socially desirable from the ministers’ viewpoint).

 

Suggested Readings

Levy, D. A. (1997). Tools of critical thinking: Metathoughts for psychology, 2^nd ed. Long Grove,

IL: Waveland Press, Inc.

This book attempts to provide the reader with a set of critical thinking tools to specifically guide problem solving in psychological research, but that could presumably be applied to other disciplines as well. It should prove particularly useful to students as it applies to discussions about course material from Chapter 6, since the book includes several chapters on causation (e.g., as being different from correlation, varying in degree, and bi-directional in nature).

 

Over, D. E., & Green, D. W. (2001). Contingency, causation, and adaptive inference.

Psychological Review, 108(3), 682-684.

The paper briefly discusses the heuristics that are used to differentially weight cells in making judgments from contingency tables, and shows how such heuristics do not always lead to appropriate conclusions. Students should gain an appreciation of the researcher’s enhanced role in the interpretation of contingency relative to correlation.

 

Suggested Web Sites

http://www.burns.com/wcbspurcorl.htm

This site from William C. Burns and Associates, a consulting firm in the San Francisco Bay area, provides a thorough, and sometimes entertaining, summary of some “spurious correlations”. The page will highlight for students by example that correlation does not indicate causation.

 

 

 

http://www.physics.csbsju.edu/stats/

This page, developed by physicist Tom Kirkman of the College of Saint Benedict and Saint John’s University, provides detailed information relating to graphical displays and the calculation of common statistics. Material relevant to Chapter 6 includes construction of contingency tables and the calculation of c², as well as least square regression analyses. In addition to descriptive information, students may find use for the included on-line statistical calculators and graphical plotters.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

TESTBANK

 

MULTIPLE CHOICE

 

1. Ex post facto data:

a.	arise from experimentation
b.	are after the fact assessments
c.	are similar to unobtrusive data
d.	only arise from a reliance on contingency tables

 

 

ANS:  B                    PTS:   1                    REF:   Introduction    MSC:  WWW

 

2. How many cells are in a contingency table that examines the frequencies of men and women who are freshman, sophomores, or juniors?

a.	2
b.	4
c.	6
d.	8

 

 

ANS:  C                    PTS:   1                    REF:   Contingency Research

 

3. A contingency table that had 4 columns and three rows would be labeled:

a.	2 ´ 3
b.	3 ´ 2
c.	3 ´ 4
d.	4 ´ 3

 

 

ANS:  C                    PTS:   1                    REF:   Contingency Research

 

4. The typical cell entries in a contingency table are:

a.	frequencies
b.	durations
c.	magnitudes
d.	ratios

 

 

ANS:  A                    PTS:   1                    REF:   Contingency Research

 

5. Which variable are you most likely to see in a relational study with human participants rather than directly manipulated in an experiment?

a.	task difficulty
b.	intensity of a light
c.	brain damage
d.	supposed level of an electric shock (via deception)

 

 

ANS:  C                    PTS:   1                    REF:   Contingency Research

 

 

 

 

6. The  c² statistic most often is used to evaluate data from ____.

a.	experiments
b.	correlational designs
c.	contingency tables
d.	observational studies

 

 

ANS:  C                    PTS:   1                    REF:   Contingency Research

 

7. In research involving after the fact assessment of the data:

a.	subject (participant) reactivity is not a problem
b.	the magnitude of reactivity is unknown
c.	frequencies are not subject to reactivity
d.	reactivity is greater than in other kinds of research

 

 

ANS:  B                    PTS:   1                    REF:   Contingency Research

 

8. The typical correlation coefficient varies from:

a.	0 to +1.00
b.	-1.0 to 0
c.	-1.0 to +1.0
d.	-10.0 to +10.0

 

 

ANS:  C                    PTS:   1                    REF:   Correlational Research

MSC:  WWW

 

9. Which of the following is an example of a positive correlation?

a.	height increases as weight decreases
b.	weight decreases as height decreases
c.	height decreases as weight increases
d.	weight stays the same as height increases

 

 

ANS:  B                    PTS:   1                    REF:   Correlational Research

 

10. Which of the following is an example of a negative correlation?

a.	height increases as weight decreases
b.	weight decreases as height decreases
c.	height increases as weight increases
d.	weight stays the same as height increases

 

 

ANS:  A                    PTS:   1                    REF:   Correlational Research

 

 

 

 

 

 

 

11. Which correlation coefficient reflects the strongest relationship between variables AND meets the following condition: an increase in the value along one variable reflects an increase in the value along the other variable?

a.	-2
b.	-1
c.	0
d.	+1
e.	+2

 

 

ANS:  D                    PTS:   1                    REF:   Correlational Research

 

12. A researcher computing the r between age and memory span might:

a.	be able to show that old age produces a decrease in memory span
b.	be able to predict memory span based on age
c.	make an error because age and memory span are on different scales
d.	be able to show the effects of age on memory span

 

 

ANS:  B                    PTS:   1                    REF:   Correlational Research

 

13. A researcher finds an r = 0. He/she might be able to conclude that:

a.	there is a strong correlation between the variables.
b.	there is a negative correlation between the variables.
c.	there is a positive correlation between the variables.
d.	there is no correlation between the variables.

 

 

ANS:  D                    PTS:   1                    REF:   Correlational Research

 

14. Which of the following conclusions can be reached given a Pearson’s r of -.9 between two variables?

a.	The value along one variable decreases with increases along the other variable.
b.	There is a very weak relationship between the two variables.
c.	There is a perfect relationship between the two variables.
d.	The direction of the relationship cannot be determined.

 

 

ANS:  A                    PTS:   1                    REF:   Correlational Research

MSC:  WWW

 

 

 

 

 

 

 

 

 

 

 

 

 

15. Which is the best estimate of Pearson’s r for the scatter plot below?

 

 

a.	2
b.	1
c.	.5
d.	0
e.	-1

 

 

ANS:  B                    PTS:   1                    REF:   Correlational Research

 

16. Probably the most common problem associated with relational research is that:

a.	researchers tend to be fallible in their observations.
b.	it does not prove causation.
c.	such studies are always performed in an artificial (laboratory) setting.
d.	there are always strong demand characteristics from the testing environment.
e.	it is impossible to replicate results.

 

 

ANS:  B                    PTS:   1                    REF:   Correlational Research

 

17. Which of the following represents a limitation (i.e., a threat to internal validity) for any correlational study of the relationship between smoking and lung cancer?

a.	Some relevant factors could not be examined in a correlational study, but all these factors could be directly manipulated in an experiment.
b.	A correlational study about smoking is more difficult to set up than a controlled experiment.
c.	Smoking incidence could be confounded with other variables that might increase susceptibility to cancer.
d.	It is not possible to construct a contingency table that maps out the relationship between variables.

 

 

ANS:  C                    PTS:   1                    REF:   Correlational Research

 

18. In the absence of additional information, it is likely that the failure to find a correlation between height and another variable for a sample of professional basketball players indicates:

a.	a violation of the statistical test.
b.	reactivity.
c.	the reliance on a truncated range.
d.	the presence of floor effects.

 

 

ANS:  C                    PTS:   1                    REF:   Correlational Research

 

 

19. Which interpretation of a zero correlation is appropriate?

a.	The range of values was too small to show a meaningful relationship between variables.
b.	There is a meaningful relationship between variables.
c.	The relationship between variables is linear.
d.	There was a lack of reactivity by the participants to the testing environment.

 

 

ANS:  A                    PTS:   1                    REF:   Correlational Research

 

20. With a truncated range, the researcher:

a.	may obtain a very low correlation.
b.	will have a broad range of scores.
c.	should correctly assume that there is no relationship between variables.
d.	may conclude that changes along one variable caused changes along the other variable.

 

 

ANS:  A                    PTS:   1                    REF:   Correlational Research

MSC:  WWW

 

21. Which conditions below should be expected to produce the lowest correlation coefficient due to a truncated range?

a.	Studying the relationship between memory and age for seniors in assisted living situations
b.	Measuring the relationship between cigarette smoking and lung cancer
c.	Examining the relationship between the teenage viewing of sexual content on television and subsequent intercourse
d.	Surveying students about sexual and religious attitudes, as well as drug use

 

 

ANS:  A                    PTS:   1                    REF:   Correlational Research

 

22. Which of the following represents an example of a truncated range?

a.	Admissions committees often restrict comparison of SAT scores to those between 500 and 800.
b.	There does not appear to be a linear relationship between age and memory.
c.	Obtained scores on a research methods exam varied from 10 to 100 percent correct.
d.	There appears to be a correlation between head size and IQ.

 

 

ANS:  A                    PTS:   1                    REF:   Correlational Research

 

 

 

 

 

 

 

 

 

 

 

 

 

23. Which value of Pearson r is closest to what you would expect to find given the curvilinear pattern of data that is displayed below?

 

 

a.	-1
b.	-.75
c.	0
d.	1

 

 

ANS:  C                    PTS:   1                    REF:   Correlational Research

 

24. Teenage sex has been shown to be more frequent following exposure to sexual content on television. However, sensation seeking behaviors also have been shown to predict teenage sex. In this description sensation seeking can most accurately be labeled as a ____.

a.	multiple regression
b.	cross-lagged panel procedure
c.	covariate
d.	causal variable

 

 

ANS:  C                    PTS:   1                    REF:   Correlational Research

 

25. Of the following options, which is most likely to improve the internal validity of a correlational research project?

a.	Replicate the study.
b.	Expand the range of a variable.
c.	Exclude possible covariates in multiple regression procedures.
d.	Use a cross-lagged panel procedure.

 

 

ANS:  D                    PTS:   1                    REF:   Correlational Research

 

26. A researcher is interested in mapping the direction and the degree of a relationship between two variables that cannot be directly manipulated. Additionally, an extended period of time (i.e., several observations) may be required before a meaningful relationship can be seen. A(n) ____ should be the most effective method of relational research for assessing this relationship.

a.	contingency table
b.	naturalistic observation
c.	cross-lagged panel procedure
d.	experiment
e.	correlation

 

 

ANS:  C                    PTS:   1                    REF:   Correlational Research

 

27. A crucial assumption underlying the cross-lagged-panel correlation procedure is that:

a.	correlations almost always indicate causation.
b.	causes are usually easy to determine from correlations.
c.	causes take time to have effects.
d.	ultimate causes derive from correlations.

 

 

ANS:  C                    PTS:   1                    REF:   Correlational Research

MSC:  WWW

 

28. Which statement about experiments is true?

a.	All experiments permit causal conclusions to be made.
b.	Variables in experiments cannot be confounded.
c.	Experiments can specify the direction, but not the magnitude, of a relationship between variables.
d.	Experiments generally offer the researcher more control than relational research.

 

 

ANS:  D                    PTS:   1                    REF:   Correlational Research

 

29. An advantage of experimentation over descriptive and relational methods is that experiments:

a.	have greater construct validity
b.	have greater external validity
c.	have lesser truncated ranges
d.	have greater internal validity

 

 

ANS:  D                    PTS:   1                    REF:   Correlational Research

 

30. Which assertion represents an ultimate cause?

a.	Memory recall was found to increase with the use of mental imagery.
b.	Evolution is responsible for our ability to speak a language.
c.	Changes in the availability of natural lighting directly affected participant estimates of the length of a day, and thus circadian rhythms.
d.	Increasing the intensity of the sound directly made it appear to last longer.

 

 

ANS:  B                    PTS:   1                    REF:   Correlational Research

 

 

TRUE/FALSE

 

1. Ex post facto data usually do not result from direct manipulation by the researcher.

 

ANS:  T                    PTS:   1                    REF:   Introduction

 

2. In a 4 ´ 2 contingency table, there are more columns than rows.

 

ANS:  F                    PTS:   1                    REF:   Contingency Research

 

3. In a 4 ´ 2 contingency table, there are 8 cells.

 

ANS:  T                    PTS:   1                    REF:   Contingency Research

 

4. Animal-rights activists claim that psychological research is more harmful to animals than is medical research.

 

ANS:  T                    PTS:   1                    REF:   Contingency Research

 

5. Contingency research is usually not done when people can appear in more than one cell of the table.

 

ANS:  T                    PTS:   1                    REF:   Contingency Research

 

6. Because it is after the fact, ex post facto research is free from subject (participant) reactivity.

 

ANS:  F                    PTS:   1                    REF:   Contingency Research

 

7. In a negative correlation, high scores on one variable go with low scores on the other variable.

 

ANS:  T                    PTS:   1                    REF:   Correlational Research

 

8. If lower scores on one variable go with lower scores on another, a negative correlation is a real possibility.

 

ANS:  F                    PTS:   1                    REF:   Correlational Research

 

9. Because there is a documented negative correlation between the number of cigarettes smoked and grades in college, we can conclude that smoking causes poor grades.

 

ANS:  F                    PTS:   1                    REF:   Correlational Research

 

 

10. Correlations are likely to be spuriously high when one or both of the variables suffer from a truncated range.

 

ANS:  F                    PTS:   1                    REF:   Correlational Research

 

 

11. A correlation coefficient of zero can be obtained even when there is actually a meaningful relationship between two variables (e.g., data is collected when there is only slight variation in one of the variables).

 

ANS:  T                    PTS:   1                    REF:   Correlational Research

 

12. A multiple regression procedure is frequently conducted when a researcher desires to predict values along a target variable/dimension using several other variables (e.g., predicting the number of words recalled from a list following using combined measures of a participant’s age and intelligence).

 

ANS:  T                    PTS:   1                    REF:   Correlational Research

 

13. Some scientists argue that experiments provide nothing more than controlled correlations between variables.

 

ANS:  T                    PTS:   1                    REF:   Correlational Research

 

14. In an experiment concerned with the effects of hours of food deprivation on the amount eaten, a proximate cause of changes in eating might be evolution.

 

ANS:  F                    PTS:   1                    REF:   Correlational Research

 

15. In an experiment concerned with the effects of hours of food deprivation on the amount eaten, an ultimate cause of changes in eating might be evolution.

 

ANS:  T                    PTS:   1                    REF:   Correlational Research

 

SHORT ANSWER

 

1. Discuss causation. Your answer should indicate the preferred scientific way of demonstrating causation.

 

ANS:

Answer not provided.

 

PTS:   1

 

2. Give examples of positive, negative, and null correlations. Draw an approximate scatter diagram of each.

 

ANS:

Answer not provided.

 

PTS:   1

 

 

 

 

 

3. Why is relative frequency (%) rather than raw frequency often a better way to understand the data in a contingency table?

 

ANS:

Answer not provided.

 

PTS:   1

 

4. Why is it difficult to determine causation from correlational research?

 

ANS:

Answer not provided.

 

PTS:   1

 

5. Suppose Plous concluded that being an animal-rights activist caused a person to have certain attitudes about medical and psychological research. What are some potential pitfalls in drawing such a conclusion?

 

ANS:

Answer not provided.

 

PTS:   1