Home
Subjects
Create
Search
Log in
Sign up
Upgrade to remove ads
Only $2.99/month
Psychology 12: chapter 8-10
STUDY
Flashcards
Learn
Write
Spell
Test
PLAY
Match
Gravity
Terms in this set (98)
bivariate correlation
definition: A statistical relationship between exactly two variables
- MEASURE two variables (either through self-report, observation, or physiological measure) in the SAME GROUP of people.
- Most appropriate when both measures are CONTINUOUS (ordinal, interval, or ratio).
correlational study
one in which all variables are MEASURED and none are manipulated
casual claim (experiment)
if a variable is MANIPULATED
statistical validity of correlation:
5 questions
1) effect size (the r value)
2) statistical significance (p value)
3) outliers
4) restriction of range
5) curvilinear association
effect size
the strength of a relationship
- in scatterplot, the tighter the line, the stronger the effect
- r ranges from -1 to 1
-directions: positive, negative, zero
strengths of r
.1 = small/ weak
.3 - medium/ moderate
.5= large/ strong
strong correlation means...
a) strong prediction
b) weak prediction
a) strong prediction
large effects are usually _____ important and small effects can still be ________.
- more
- important
statistical significance (p-value)
the likelihood that the effect would come out that strong by chance, assuming no effect in the real world
-p value less than 0.05 = significant
- p value greater than 0.05 = not significant
P-values are influenced by
larger effects and larger samples
- causing a smaller p value
outliers
an extreme score on either or both variables
outliers are a problem because....
1) can exert a strong, disproportionate influence on the size of the correlation
2) can make correlation much stronger/ weaker than it would have been
3) can make the correlation significant/ not significant
outliers can have a _______ effect on small samples
greater
restriction of range
when you only have part of the full scale for one or more of your variables represented in the sample
- artificially reduces the strength of the correlation
example: GPA scale when it is 3.0 to 2.0
curvilinear relationship
A relationship that is not well-represented by a straight line (not linear) because normal correlation (r) only tests for LINEAR relations
- correlations will be weak
requirements to determine causation:
1) covariance of cause and effect: as one changes, the other also changes
2) temporal precedence: cause must precede effect; it must come first in time
3) internal validity: nothing else can explain the relationship between the two variables
- directionality problem: we don't know which variable comes first
- third variable problem: failing on #3- meaning there's a third variable to consider
Suppose you hear that conscientious people are more likely to get regular health checkups. Which of the following correlations between conscientiousness and getting checkups would probably support this claim?
a. r = .03
b. r = .45
c. r = −.35
d. r = −1.0
b. r = .45
Which of these associations will probably be plotted as a bar graph rather than a scatterplot?
a. The more conscientious people are, the greater the likelihood they'll get regular health checkups.
b. Level of depression is linked to the amount of exercise people get.
c. Students at private colleges get higher GPAs than those at public colleges.
d. Level of chronic stomach pain in kids is linked to later anxiety as adults.
c. Students at private colleges get higher GPAs than those at public colleges.
A study found that people who like spicy foods are generally risk takers. Which of the following questions interrogates the construct validity of this correlation?
a. Is the result statistically significant?
b. Did the study use a random sample of people?
c. Were there any outliers in the relationship?
d. How well did they measure each variable, risk taking and liking spicy foods?
d. How well did they measure each variable, risk taking and liking spicy foods?
Darrin reads a story reporting that students at private colleges get higher GPAs than those at public colleges. He wonders if this means going to a private college causes you to have a higher GPA; if so, he'll go to a private college! Applying the three causal criteria, Darrin knows there is covariance here. He also knows there is temporal precedence because you choose a college first, and then you get your GPA. Which of the following questions would help Darrin ask about the third criterion, internal validity?
a. Could there be restriction of range?
b. Is the link between private college and high grades the same for both men and women?
c. How did they decide what qualifies a college as private or public?
d. Is there some other reason these two are related? Maybe better students are more likely to go to private colleges, and they are also going to get better grades
d. Is there some other reason these two are related? Maybe better students are more likely to go to private colleges, and they are also going to get better grades
Which of the following sentences describes a moderator for the relationship between risk taking and liking spicy foods?
a. There is a positive relationship between liking spicy foods and risk taking for men, but no relationship for women.
b. Older adults tend to like spicy foods less than younger adults.
c. The relationship between liking spicy foods and risk taking is the same for people in cities and in rural areas.
a. There is a positive relationship between liking spicy foods and risk taking for men, but no relationship for women.
correlation does not imply causation, but sometimes, correlation is the only option available to researchers
-impossible to manipulate age, preferences, and many other types of variables
- some types of experimental manipulations would be unethical to perform
multivariate designs
correlational studies that involve more than two variables
two approaches with multivariate designs
1) longitudinal designs: measuring the same variable(s) repeatedly at several points in time
2) multiple- regression analyses: predicting an outcome from more than one predictor variable to narrow down the relationship with the predictor of interest
longitudinal design
example: measuring optimism and anxiety at 9 points in time
- before one took the test, then eight times during the waiting period
correlations in longitudinal designs
cross- sectional correlation: are the two variables correlated within the same point in time? ( just like measuring them at one point in time)
auto correlations: is each variable related to itself across time measured on two occassions?
cross- lag correlations: is the earlier measure of one variable associated with the later measure of the other variable?
- address the directionality problem and helps establish temporal precedence
- if one cross lag is stronger than the other, suggests that one variable precedes the other
if both cross lag is strong (or weak), there is no temporal precedence
univariate regression
is like correlation
- asks about variance in CRITERION variable can be accounted for by the PREDICTOR variable
explaining variance
strong correlation/ regressions are said to explain/ account for the variance in the other variable
logic of multiple regression
what if we just measured every potential confound and statistically CONTROLLED for their effects on the outcome?
- it would solve internal validity
in reality, it's not possible or feasible to measure EVERY confound, but multiple regression helps rule out some confounds
- multiple regression will tells us how much variance each variable uniquely accounts for
- it tells us the association of one predictor while CONTROLLING for the other
statistically controlling for other predictors
what is the association between ONE PREDICTOR and the CRITERION while controlling for the other PREDICTORS
- hold the other predictors constant to look for the association between one predictor and the criterion
controlling for effects of other variables
CONTROLLING: eliminating the effects of a third variable to look at the unique contribution of the predictor
- hold one variable constant to see the effects of another
multiple regression analysis
multiple regression analysis attempt to rule out third variables (increasing internal validity)
controlling for: holding a potential third variable at a constant level while investing at the association between two other variables
- if both variables are related, then rules out third variable
-if both variables are not related, then undermines hypothesis that iv1 leads to iv2
criterion variable
dependent variable- the one you are trying to predict
predictor variable
independent variable- variable that might be causing the change in the dependent variable
beta
statistical representation of the relationship between each predictor and criterion variable
- ranges from -1 to 1
- has an associated p value
- difference is that the relationship controls for the other predictor variables
beta statistic
the strength of the association between a predictor and the criterion WHILE CONTROLLING FOR THE OTHER PREDICTORS
negative= negative association
positive= positive association
zero= no association
beta statistic indicates...
a relatively stronger association for one predictor compared to others
multiple regression analysis CANNOT
establish causation because...
1) temporal precedence could still be in question
2) might not have measured the right third variable
Mediation
tells us why two variables are associated
- when, or for whom, are two variables associated
mediator
a variable that explains the relationship between two other variables
- answers they why question
- testing mediation involves correlations and multiple- regression
Four steps to establish mediation
1) test relationship A to C (called path c)
2) test relationship A to B (called path a)
3) test relationship B to C (called path b)
4) multiple- regression" does c hold up rather controlling for the mediator? (c')
logic of mediation
- if we have a relation between A and C, is there a variable B that is responsible for that association?
- if we know that A and C are related, we say that B is a mediator if A predicts B and B predicts C
- mediator are internal to the relationship of interest while confounds are external
- third variables are a nuisance that we want to control for, but mediators are of director interest
confound
can a third variable explain why two variables are related
mediators vs confounds
a mediator explains why two things are causally related, whereas a third variable explains why two things seem causally related but really aren't
A headline in Yahoo! News made the following (bivariate) association claim: "Facebook users get worse grades in college" (Hsu, 2009). The two variables in this headline are:
a. Level of Facebook use and college grades.
b. High grades and low grades.
c. High Facebook use and low Facebook use
a. Level of Facebook use and college grades.
Suppose a researcher uses a longitudinal design to study the relationship between Facebook use and grades over time. She measures both of these variables in Year 1, and then measures both variables again in Year 2. Which of the following is an example of an autocorrelation in the results?
a. The correlation between Facebook use in Year 1 and Facebook use in Year 2.
b. The correlation between Facebook use in Year 1 and grades in Year 2.
c. The correlation between grades in Year 1 and Facebook use in Year 2.
d. The correlation between grades in Year 1 and Facebook use in Year 1.
A) autocorrelation is the same variable at different times.cross-sectional is different variable (hence 'cross'), same time.cross-lag is IV-->DV andDV--> IV
In the longitudinal study described in question 2, which pattern of cross-lag correlations would indicate that Facebook use leads to lower grades (rather than the reverse)?
a. Grades at Year 1 shows a strong correlation with Facebook use at Year 2, but Facebook use at Year 1 shows a weak correlation with grades at Year 2.
b. Grades at Year 1 shows a weak correlation with Facebook use at Year 2, but Facebook use at Year 1 shows a strong correlation with grades at Year 2.
c. Grades at Year 1 shows a strong correlation with Facebook use at Year 2, and Facebook use at Year 1 shows a strong correlation with grades at Year 2.
B)grades (dv) at year 1 weak correlation to Facebook use (iv) at year 2 &&Facebook use (iv) at year 1 strong correlation to grades (dv) at year 2.
Consider this statement: "People who use Facebook got worse grades in college, even when the researchers controlled for the level of college preparation (operationalized by SAT scores) of the students." What does it mean?
a. Facebook use and grades are correlated only because both of these are associated with SAT score.
b. SAT score is a third variable that seems to explain the association between Facebook use and grades.
c. SAT score can be ruled out as a third variable explanation for the correlation between Facebook use and college grades.
C) SAT scores are not a third variable because the association remained
Which of the following statements is an example of a mediator of the relationship between Facebook use and college grades?
a. Facebook use and college grades are more strongly correlated among nonathletes, and less strongly correlated among athletes.
b. Facebook use and college grades are only correlated with each other because they are both related to the difficulty of the major. Students in more difficult majors get worse grades, and those in difficult majors have less time to use Facebook.
c. Facebook use and college grades are correlated because Facebook use leads to less time studying, which leads to lower grades
C) answers question of why is Facebook use correlated to grades
A news outlet reported on a study of people with dementia. The study found that among patients with dementia, bilingual people had been diagnosed 3-4 years later than those who were monolingual. What are the variables in this bivariate association?
a. Being bilingual or monolingual
b. Being bilingual or not, and age at dementia diagnosis c. Age at dementia diagnosis
B) bilingualism and age of diagnosis
The journalist reported that the relationship between bilingualism and age at diagnosis did not change, even when the researchers controlled for level of education. What does this suggest?
a. That the relationship between bilingualism and dementia onset is probably attributable to the third variable: level of education.
b. That the relationship between bilingualism and dementia onset is not attributable to the third variable: level of education.
c. That being bilingual can prevent dementia.
B) education is not a third variable; the association between bilingualism and age at diagnosis remains even after accounting for this.
Researchers speculated that the reason bilingualism is associated with later onset of dementia is that bilingual people develop richer connections in the brain through their experiences in managing two languages; these connections help stave off dementia symptoms. This statement describes:
a. A mediator
b. A moderator
c. A third variable
a. A mediator
a new study reports that multiple head injuries predict memory impairment, even after controlling for level of education and severity of injuries.Here the dependent variable, also called the __ variable, is___.
A) independent; head injuries
B) criterion; memory impairment
C) criterion; head injuries
D) covariate; emory impairment
C) criterion variable means the same thing as dependent variable; memory impairment
when researchers conduct longitudinal research they are most interested in which of the following correlations?
A) autocorrelation
B) cross-sectional correlation
C) cross-lag correlation
D) all bivariate correlations
C) cross-lag correlations
autocorrelation; same idea, different times.cross-sectional; different ideas, same timecross lag; IV early, DV later and DV early to IV late.
All of the following are true of betas and correlation coefficients except:
A) betas are the same thing as correlationsB
) both betas and correlation coefficients can tell you something about the strength of a relationship
C) both betas and correlation coefficients can tell you something about the direction of a relationship
D) standardized betas from an analysis can be compared with other standardized beta co efficient from the same analysis just as correlation coefficients can
A) they are not the same thing but they do have similarities
difference between third variable and mediator
third variable: comes before both the IV and the DV and makes the relationship between them spurious.mediator: explains the relationship between the IV and the DV
having a cognitively demanding job is associated with cognitive benefits in later years because people who are highly educated take cognitively demanding jobs, and people who are highly educated have better cognitive skills.
A) third variable
B) moderator
C) mediator
D) none
A) third variable
viewing violent television is associated with aggressive behavior because children model what they see on TV.what type of prediction is this?
A) third variable
B) moderator
C) mediator
D) none of the above
C) mediator
children acting out what they see on TV is why the TV makes them more aggressive.
viewing violent television is associated with aggressive behavior because people who watch more violent TV have more lenient parents, and these lenient parents also do not care is their children are violent.which type of prediction is this?
A) third variable
B) moderator
C) mediator
D) none of the above
A) third variable. lenient parents accounts for each of the other variables.
he Yerkes-Dodson law (1908), shows that performance increases with arousal up to a point, but beyond that, performance decreases with increasing arousal. What type of relationship is this?
A) curvilinear
B) zero
C) positive
D) negative
A) curvilinear
Which of the following is NOT a way that a researcher might indicate a statistically significant result in a journal article?
A) asterisk *
B) notation of P<.05
C) notation of p = .20
D) the word sig
C) notation of p = .20
When is an outlier most likely to be problematic?
A) when the sample size is large and the outlier is extreme on both variables
B) when the sample size is small and the outlier is extreme on both variables
C) when the sample size is large and the outlier is extreme on one variable
D) when the sample size is small and the outlier is extreme on one variable
B) small sample size and extreme on both variables
which of the following is true about the difference between beta and r?
A) unlike r, beta of zero reflects a statistically significant relationship between the predictor variable and the criterion variable
B) unlike r, the size of beta is not related to the magnitude of the effect
C) unlike r, beta cannot be negative
D) unlike r, beta reflects the independent contribution of the predictor variable, controlling for the contributions from the other predictor variables
D) beta reflects the contribution of the IV, controlling for the impact of other IVs. (third variables)
What type of research design involves measuring the same variables, for the same people, across different points in time?
A) multiple regression
B) cross-sectional
C) longitudinal
D) pattern and parsimony
C) longitudinal
Dr. Russell did a study that found that praise provided by supervisors is associated with higher levels of work productivity only because more motivated employees are praised more often, and highly motivated people are more productive. In her findings, employee motivation is a ___________ in the relationship between praise from supervisors and work productivity.
A) moderator
B) criterion variable
C) mediator
D) confounding third variable
D) confounding third variable
How do multiple-regression designs help address internal validity?
A) establishing temporal precedence
B) introducing a control condition
C) eliminating selection threats
D) trying to rule out or account for 3rd variables
D--third variables
control variables
Any variable the researcher intentionally holds constant across conditions
independent is to _______ variable as dependent variable is to _____ variable
-manipulated
-measured
conditions
the levels or versions of the independent variable
Two ways to check construct validity of manipulated variables
-pilot study: conducted before the actual study to check the construct validity of a manipulation
-manipulation check: an extra measure designed to see how well a manipulation worked
types of conditions
- control group: represent no treatment group
- treatment group
placebo group: group believes they're a treatment group, with the goal of ruling out expectancy effect
internal validity in experiments
A well-designed experiment maximizes internal validity, compared to other designs
- confound: anything that differs between your group OTHER THAN the levels of the independent variable
- design confound: something that inherently varies along with the independent variable
- creates systematic variability (a problem)
- unsystematic variability: created when something differs among participants but does NOT systematically co-occur with the independent variable
a TRUE EXPERIMENT has two essential characteristics:
- manipulation of one or more independent variables
- random assignment to conditions
random assignment
each participant has an equal chance of being placed into any group/ condition
-maximizes the likelihoof thst unintended variability is unsystematic instead of systematic
-avoids SELECTION EFFECTS: when the kind of person in one condition are systematically different from the ones in the other conditions
failure of random assignment
can happen when randomness doesnt seem random
matched groups or matching
Ensuring that your groups are equivalent in important ways
Pair (match) people on the characteristic of interest, then split the pair across conditions through random assignment
between subjects design
each participant experiences only one level of the independent variable
Two types of between-subjects designs
-posttest only design: participants undergo the manipulation (just one condition) and then complete the measures
-pretest-posttest design: participants first complete the measures, then the manipulation, then the measures again
advantages and disadvantages of pretest-posttest design
advantages:
- test and control for selection effects
-test and control for failures of random assignment
disadvantages:
-might create DEMAND CHARACTERISTCS
- people might think they should be consistent in their responses
within subject design
each participant is in all experimental conditions
two types of within subjects design
-CONCURRENT MEASURES: participants experience all levels of the independent variable at once
- REPEATED MEASURES: participants experience levels of the independent variable one after the other, with the measures following each level of the independent variable
why would you want a within subject design
- guarantees equivalence of groups
- functionally doubles your sample size (for two conditions)
statistical power
Ability of a study to get a statistically significant effect, assuming the effect is real
why would you want a between subjects design
order effects- a confound that occurs when experiencing one condition changes how participants react to subsequent conditions in a within subject design
types of order effects
(1) Practice effects
(2) Fatigue effects
(3) Carryover effects
(4) Sensitization effects
practice effect
Participants may get better at the task because of repeated exposure (rather than development)
fatigue effect
participants perform a task worse in later conditions because they become tired or bored
carryover effects
Effects of one condition contaminate subsequent responses
sensitization effect
participants become suspicious or clued in from earlier conditions
how do you deal with some order effects
counterbalancing
Counterbalancing
randomly assigning participants to experience the conditions in different orders
types of counterbalancing
-FULL: all possible orders are represented
-PARTIAL: only some orders are represented
TRUE or FALSE:
counterbalancing DOESNT FIX order effects, it allows you to check for them
TRUE
Max ran an experiment in which he asked people to shake hands with an experimenter (played by a female friend) and rate the experimenter's friendliness using a self-report measure. The experimenter was always the same person, and used the same standard greeting for all participants. People were randomly assigned to shake hands with her either after she had cooled her hands under cold water or after she had warmed her hands under warm water. Max's results found that people rated the experimenter as more friendly when her hands were warm than when they were cold.
Why does Max's experiment satisfy the causal criterion of temporal precedence?
a. Because Max found a difference in rated friendliness between the two conditions, cold hands and warm hands.
b. Because the participants shook the experimenter's hand before rating her friendliness.
c. Because the experimenter acted the same in all conditions, except having cold or warm hands.
d. Because Max randomly assigned people to the warm hands or cold hands condition.
b. Because the participants shook the experimenters' hand before rating her friendliness
Max ran an experiment in which he asked people to shake hands with an experimenter (played by a female friend) and rate the experimenter's friendliness using a self-report measure. The experimenter was always the same person, and used the same standard greeting for all participants. People were randomly assigned to shake hands with her either after she had cooled her hands under cold water or after she had warmed her hands under warm water. Max's results found that people rated the experimenter as more friendly when her hands were warm than when they were cold.
In Max's experiment, what was a control variable?
a. The participants' rating of the friendliness of the experimenter.
b. The temperature of the experimenter's hands (warm or cold).
c. The gender of the students in the study.
d. The standard greeting the experimenter used while shaking hands.
d. The standard greeting the experimenter used while shaking hands
Max ran an experiment in which he asked people to shake hands with an experimenter (played by a female friend) and rate the experimenter's friendliness using a self-report measure. The experimenter was always the same person, and used the same standard greeting for all participants. People were randomly assigned to shake hands with her either after she had cooled her hands under cold water or after she had warmed her hands under warm water. Max's results found that people rated the experimenter as more friendly when her hands were warm than when they were cold.
What type of design is Max's experiment?
a. Posttest-only design
b. Pretest/posttest design
c. Concurrent-measures design
d. Repeated-measures design
a. Posttest-only design
Max ran an experiment in which he asked people to shake hands with an experimenter (played by a female friend) and rate the experimenter's friendliness using a self-report measure. The experimenter was always the same person, and used the same standard greeting for all participants. People were randomly assigned to shake hands with her either after she had cooled her hands under cold water or after she had warmed her hands under warm water. Max's results found that people rated the experimenter as more friendly when her hands were warm than when they were cold.
Max randomly assigned people to shake hands either with the "warm hands" experimenter or the "cold hands" experimenter. Why did he randomly assign participants?
a. Because he had a within-groups design.
b. Because he wanted to avoid selection effects.
c. Because he wanted to avoid an order effect.
d. Because he wanted to generalize the results to the population of students at his university.
b. Because he wanted to avoid selection effects
Max ran an experiment in which he asked people to shake hands with an experimenter (played by a female friend) and rate the experimenter's friendliness using a self-report measure. The experimenter was always the same person, and used the same standard greeting for all participants. People were randomly assigned to shake hands with her either after she had cooled her hands under cold water or after she had warmed her hands under warm water. Max's results found that people rated the experimenter as more friendly when her hands were warm than when they were cold.
Which of the following questions would be interrogating the construct validity of Max's experiment?
a. How large is the effect size comparing the rated friendliness of the warm hands and cold hands conditions?
b. How well did Max's "experimenter friendliness" rating capture participants' actual impressions of the experimenter?
c. Were there any confounds in the experiment?
d. Can we generalize the results from Max's friend to other experimenters with whom people might shake hands?
b. how well did Max's "experimenter friendliness" rating capture participants' actual impressions of the experimenter?
Other sets by this creator
psych 110: vision system to motor
12 terms
psychology 12 final
214 terms
Psychology 12: Chapter 1-3
51 terms
starbucks shots and syrups
53 terms