PSY 400 - EXAM 2 (7, 8, 10, 11, 13, 14)

___________ controls for order effects in within-groups designs. a. counterbalancing b. using a manipulation check c. conducting a pilot study d. making unobtrusive observations

counterbalancing (a)

Autumn criticized my experiment for its terrible internal validity. Which of these did she probably say? a. "You didn't use a within-groups design." b. "You didn't use random assignment." c. "You didn't use a random sample." d. "You didn't use a pretest-posttest design."

"You didn't use random assignment." (b)

Define independent variable, dependent variable, and control variable.

- Independent variable - manipulated variable - Dependent variable - measured (outcome) variable - Control variable - variable they hold constant and control for

What are four ways of selecting a nonprobability sample?

1. Convenience sampling 2. Purposive sampling 3. Snowball sampling 4. Quota sampling

What are the three types of replication studies?

1. Direct 2. Conceptual 3. Replication-plus-extension

What are six questions you can ask about the statistical validity of a bivariate correlation?

1. How strong is the relationship? (strength) 2. How precise is the estimate? (Precision) 3. Has it been replicated? (replicability) 4. Could outliers be effecting your association? (outliers) 5. Is there restriction of range? 6. Is the association curvilinear?

What are five sampling techniques for selecting a probability sample of a population of interest?

1. Simple random sampling 2. Cluster sampling 3. Systematic sampling 4. Stratified random sampling 5. Oversampling

What are three reasons a researcher might conduct a quasi-experiment, rather than a true experiment, to study a research question?

1. real-world opportunities 2. ethics 3. want to maximize external validity, so people who are related to the study are chosen

At a minimum, how many variables are there in an association claim?

2

Using people who are readily available to the researcher for a study may lead to:

A biased sample

Which of these samples is more likely to be representative of a population of 100,000? a. a snowball sample of 50,000 people b. a cluster sample of 50,000 people

A cluster sample of 50,000 people. (snowballs can't generalize)

Which of the following is never found in a one-group, pretest/posttest design? a comparison group a manipulation a valid measure a random sample

A comparison group

RESEARCH STUDY 11.1: In previous studies, Dr. Hamid has established that finding meaning in one's everyday work activities can lead to greater success in the workplace (e.g., productivity, creativity). He is curious as to whether this can happen in the college classroom. Specifically, he is curious whether finding meaning in one's classroom experience can lead to greater academic performance. In the spring semester, he has his teaching assistant randomly assign half the class to write a paragraph each class period about how the material has meaning for their lives (meaning group). The other half writes a paragraph about what they did to prepare for class (preparation group). He does not know which of his students are writing which paragraph, and the students are not aware they are responding to different writing assignments. To measure academic performance, he gives the students a midterm essay exam and a final exam. The study described above is an example of which of the following? a one-group, pretest/posttest design a double blind study a null effect observer bias

A double blind study

A study conducted in the "real world" is often said to be conducted in:

A field setting

What can a meta-analysis tell you?

A meta-analysis tells you a statistical summary of a lot of studies

RESEARCH STUDY 11.2: Dr. Bloedorn is a health psychologist who researches nutrition. She is curious as to whether a new drink additive will help people consume fewer calories during a meal. The drink additive is a white, odorless, tasteless powder that a person can add to any drink. She collects a random sample of 63 overweight students on campus and measures the calories they eat during lunch using a bomb calorimeter. She then gives this additive to the same 63 participants to use at dinner and measures how many calories they eat (again using the bomb calorimeter). Imagine that Dr. Bloedorn finds no difference between the calories consumed with the drink additive and without. This is known as: the placebo effect a maturation effect a null effect observer bias

A null effect

Dr. Morimoto is curious as to whether exposing people to violent video games causes them to be more aggressive. He assigns half his participants to play a video game for 5 minutes and the other half to play for 7 minutes. He finds that there is no relationship between playing the game longer and being more aggressive. What might be to blame for this null effect?

A weak manipulation

What characteristics of a study makes it correlational?

All variables are measured

In one or two brief sentences, explain how you would interrogate the construct validity of a bivariate correlation.

Ask how well each variable was measured or operationalized

Random assignment and random sampling are necessary for which validities?

Assignment - necessary for internal validity Sampling - necessary for external validity

Why do experiments usually satisfy the three causal criteria (covariance, temporal precedence, internal validity)

At the end of an experiment we can see whether the two variables have a relationship or covary, (covariance) we know which variable came first, (temporal precedence) and if the research was designed well and the control variables were done well, then we know no other explanations can be made for the association (internal validity)

Explain, in your own words, what a replication project does.

Attempts to provide additional evidence for claims that have already been made

Dr. Granger notices that 20 students in their longitudinal study of 100 college students dropped out of the experiment over time. When they look at the missing data, they discover that those 20 students had significantly lower pretest scores than the 80 with complete data. Which type of threat is this an example of? history testing instrumentation attrition

Attrition

Dr. Wilson conducted a long-term study on friendship. He noticed that the most introverted people dropped out by the third session. Therefore, his study might have which of the following internal validity threats? a. Attrition b. Maturation c. Selection d. Regression

Attrition (a)

In a study design, Greg is thinking is having one group of designers hold a 30-minute conversation with the client, and the other group hold a 15-minute conversation with the client. This would probably: a. Work well because both groups are having a conversation with the client b. Be a problem because it would increase the within-groups variability c. Be a problem because there would not be enough between-groups variability

Be a problem because there would not be enough between-groups variability (c)

Which of the following four terms does not belong: generalizable sample, externally valid sample, representative sample, biased sample

Biased sample

How are quota sampling and stratified random sampling similar?

Both identify subgroups that need to be studied

"Having mom in the car makes teens safer drivers" is an example of which type of claim?

Causal

Describe why under-reporting p-hacking is a questionable research practice.

Changing the data to fit into your hypothesis makes it seem like a relationship exists where it doesn't, and misleads researchers / the public

The difference between a cluster sample and a stratified random sample is:

Cluster samples use randomly selected clusters, stratified random samples use predetermined strata

How do cluster sampling and multi-stage sampling go hand-in-hand?

Cluster sampling randomly selects clusters for a study, multi-stage sampling then randomly selects participants out of those clusters.

If researchers measure every member of a population, they have:

Conducted a census

What is the most common sampling technique in behavioral research?

Convenience sampling

Some people have a gene that allows them to detect bitter tastes, even in very small doses. A study found that people who had this bitter-tasting ability were jumpier in response to loud noise, compared to those who did not have the ability to taste bitter substances. This study is: a. Correlational b. Experimental

Correlational (a)

Describe how counterbalancing improves the internal validity of a within-groups design

Counterbalancing takes away order effects, allowing the researcher to change the order in which the questions are presented

Which of the three rules of causation is almost always met by a bivariate correlation?

Covariance

Keisha's study has a posttest-only design. Therefore, she can rule out all of the following EXCEPT: a. Regression to the mean b. Design confounds c. Maturation d. Attrition

Design confounds (b)

RESEARCH STUDY 14.1: Melanie conducts a study for her research methods class to determine if consuming caffeine causes people to perform better on cognitive tasks. In her study, she gives half of her participants a glass of water and half of her participants a glass of cola and then has them attempt to solve 15 math problems. She finds that people who drink caffeine beforehand solve more math problems than those who drink water beforehand. Melanie decides to conduct the exact same study again to ensure that she can find the same difference between the groups a second time. This study is known as a:

Direct replication

In one study, certain participants were asked to ingest some of a very bitter herbal drink. Then they were asked to judge how morally wrong several activities were (such as stealing library books or accepting a bribe). Compared to participants who were asked to drink water, those who tasted the bitter substance were more likely to judge the activities as more morally wrong. This study is: a. Correlational b. Experimental c. Quasi-experimental

Experimental (b)

For what type of claim will it be most important for a researcher to use a representative sample?

Frequency claims

When conducting quasi-experiments, researchers are forced to give up: a. Full experimental control b. Real-world applicability c. Traditional ethical standards

Full experimental control (a)

If some outside event (such as a change in weather) influences most of the people in a study between pretest and posttest, this would be a(n): a. history threat b. selection threat c. order effect d. placebo effect

History threat (a)

When you are interrogating the external validity of a sample, which is the most important question to ask?

How was the sample collected?

Natasha complains that I needed a bigger sample in my study. When would I need to listen to her? a. If my study shows a null effect b. If I need to generalize to my population of interest c. If I have a within-groups design

If my study shows a null effect (a)

What is the difference between independent-groups and within-groups designs? Use the term levels in your answer.

Independent-groups (between-subjects design) - different groups are assigned to experience different levels of the independent variable. No participant experiences all levels. Within-groups (within-subjects design) - each participant experiences EVERY level of the independent variable.

Describe why under-reporting null effects is a questionable research practice.

It leaves out important data, and is misleading to other researchers

Describe why under-reporting HARKing is a questionable research practice.

It makes it seem like you knew all along and hypothesized something you didn't Can also make you sound dumb in your hypothesis if an unrelated topic is mentioned out of nowhere

When might researchers decide to use a nonprobability sample, even though a probability sample would ensure external validity?

Lack of funds or if priorities lie elsewhere

When conducting a poll, adding more people to the sample will: a. make the sample more externally valid b. make the margin of error of the estimate smaller c. both A and B are correct

Make the margin of error of the estimate smaller (b)

Increasing sample size from 500 to 1,000 does what?

Makes the margin of error smaller

23% of a sample, plus or minus 4%. What is another term for the 4% value?

Margin of error

Martin has found a correlation of r = .18 between the two variables of using prescription stimulants (e.g. Adderall) and frontal lobe activity. This correlation is more likely to be statistically significant if:

Martin used a larger number of subjects

Devon wonders, "What can I do to avoid selection effects?" You can advise him to use a. A pretest-posttest design b. matching c. a repeated measures design d. random sampling

Matching (b)

A variable the researcher controls is a _________ variable.

Measured

What is a reason that psychologists especially value meta-analysis?

Meta-analyses allow researchers to examine the strength of a relationship

How would a researcher quantitatively summarize a scientific literature? a. Review article b. Literature review c. Meta-analysis

Meta-analysis (c)

The difference between a cluster sample and a multi-stage sample is:

Multi-stage samples sample both clusters and participants, cluster samples just sample clusters.

How is a nonequivalent control group different from a true independent-groups design?

There is no random assignment in nonequivalent design

What is an advantage of studies that are conducted in real-world settings?

They are high in ecological validity

According to the textbook, what is the problem with WEIRD samples?

They are not very representative of the world's population

Why is it important for studies with null results to be published?

They push researchers to revise existing theories.

What do purposive, convenience, quota, and snowball sampling have in common?

They result in samples where some people are systematically left out

Why do researchers want a larger sample? a. to maximize external validity b. to maximize statistical validity c. to maximize construct validity d. to maximize internal validity

To maximize statistical validity (b)

While reading about a research study, what would tell you that an association claim is being made?

Two measured variables

What are the five components of WEIRD people?

W - Western E - Educated I - Industrialized R - Rich D - Democratic

What is a WEIRD sample

W - Western E - Educated I - Industrialized R - Rich D - Democratic

Does a large sample size still need to be a random sample in order to be generalizable?

Yes

Does sample size primarily effect statistical validity?

Yes

Which of the following is a method for preventing selection effects? a. posttest-only design b. matched-groups design c. pretest/posttest design

matched-groups design (b)

A (negative) correlation was found between a state's average level of extroversion and the degree of mountainousness in that state. r = -.40; 95% Ci [-.14, -.61] The results suggest that: a. more mountainous states have fewer extroverts b. more mountainous states have more extroverts

more mountainous states have fewer extroverts (a) (negative relationship = more, less positive relationship = more, more or less, less)

Why are curvilinear relationships hard to detect with correlation coefficients (r)?

r always looks for the best straight line to fit the data

The relationship on a graph was described as r = -.40; 95% CI [-.14, -.61]. Therefore,

the relationship is negative, the correlation is statistically significant, and we can probably rule out that the relationship is zero

List 3 reasons why a study could have obtained a null effect:

too much unsystematic variability, not enough variability between groups, the independent variable didn't make a difference

Which two rules of causation might not be met by a correlational study?

Temporal precedence Internal validity

A study finds a correlation coefficient of r = .52 and reports 95% CI [.37, .67]. The 95% CI indicates:

The correlation is unlikely to have come from a zero population

The degree to which a quasi-experiment supports a causal statement depends on what two things?

The quality of the design and the pattern of results

If all the students in the two classes that are offered a survey take it, what has happened?

The researcher relied on a census

A study finds a correlation of r = .52. This number gives you information about what?

The strength and direction of the relationship

You can tell when a study is correlational because: a. the variables are all measured. b. the researchers used the correlation coefficient, r. c. the researchers used a t-test. d. the researchers used a scatterplot.

The variables are all measured (a)

What are the minimum requirements for a study to be an experiment?

There has to be a manipulated variable, and a measured variable.

3 participants in a study that seem very extreme and nowhere near the other points on the scatterplot should be considered _______?

Outliers

Why is there a publication bias against null effects?

People generally want to read about independent variables that change the outcome

Greg's dependent measure was the effectiveness of the design (of a cafeteria catered towards people in wheelchairs) as measured by a panel of expert designers. They rated the designs as 0 (not effective), 1 (effective), or 2 (very effective). This measure would: a. Work well because it only has 3 easy categories b. Possibly decrease the between-groups variability because of an insensitive measure c. Possibly cause a confound because everybody uses this scale differently

Possibly decrease the between-groups variability because of an insensitive measure (b)

In another study of hockey fans, researchers approached men who had just cheered for the winning team as well as those who had just cheering for the losing team and asked them to taste and rate a citrus-flavored sorbet. The men whose team had just won rated the sorbet as sweeter (and less sour) compared to the losing team's fans. This study is: a. Correlational b. Experimental c. Quasi-Experimental

Quasi-experimental (c)

Obtaining particular participants by posting a flyer in a local grocery store, is an example of what sampling technique?

Quota sampling

How does random assignment prevent selection effects?

Randomly assigning people to groups makes groups equal to each other

The first step in establishing a study's quality and credibility is to establish its:

Replicability

Some studies may not replicate because the original study used questionable scientific practices. What is an example of a way open science practices attempt to deal with this problem?

Requiring scientists to make their data and materials available for review

Which of the following threats to internal validity would result in group differences prior to the start of the study? selection effect design confound maturation effect order effect

Selection effect

Experiments use random assignment to avoid:

Selection effects

Random assignment can be used to avoid which of the following? a. practice effects b. order effects c. selection effects d. carryover effects

Selection effects (c)

Online surveys commonly suffer from:

Self-selection

A researcher's attempt to control sounds, smells, and even temperature in a testing environment is meant to reduce which of the following? situation noise between-groups variance demand characteristics weak manipulation

Situation noise

Ceiling effects can lead to:

Small variance between groups

Asking participants for contacts of theirs in order to get more participants is an example of what kind of sampling?

Snowball sampling

Which of the four validities does the number of participants address? a. Construct b. Internal c. External d. Statistical

Statistical (d)

In your own words, describe the steps a researcher follows in a meta-analysis.

Step 1: Collect all studies on a specific topic Step 2: Summarize all of the studies statistically

Convenience sampling relies on what?

Studying people who are easy to find

According to the book, what does the E stand for in WEIRD?

educated

What does HARKing stand for?

hypothesizing after results are known