RM 2 Exam #2
A researcher wants to examine the relationship between one's popularity on Twitter, and the reach of their content. The researcher defines "popularity" as the number of individuals that each person follows, and "reach of content" as the number of times their posts gets retweeted. What is true about the proposed study?
"Popularity" as defined by the researcher has low construct validity
The partipants subsection should enumerate
-How many participants came to the lab/ started the study -how many participants withdrew from the study -how many participants were excluded
What is true about within-subjects designs?
-They're more economical -They have greater statistical power than between-subjects designs -They may involve order effects -All of the above
A 2 x 3 factorial design will have how many interactions?
1
A 2x3 factorial design will have _____ interactions
1
A 2x3 factorial design will have how many interactions?
1
In a 2 x 3 factorial design, where both independent variables are between-subjects, each participant is assigned to __ condition(s) out of ___ conditions in the experiment.
1; 6
In a 2 x 3 factorial design, where both independent variables are manipulated between-subjects, each participant is assigned to ____ out of ____ unique combinations of conditions in the experiment.
1; 6
In a 3 x 3 factorial design, where both independent variables are manipulated between-subjects, each participant is assigned to ____ out of ____ unique combinations of conditions in the experiment.
1; 9
2 x 3 factorial design will have ___ main effects
2
3 x 3 factorial design will have ___ main effects
2
A 2 x 3 factorial design will have ____ main effects.
2
How many independent variables are there in a 2x3 design?
2
A factorial design with 3 IVs has _____ main effects.
3
A factorial design with 3 independent variable has ___ main effects
3
What should NOT be part of the Procedure?
A description of a typical session from your perspective/ the perspective of the researcher
If each participant is exposed to all levels of the independent variable, the experiment is:
A within-participants design
A researcher is interested in examining whether the emotional valence of sentences affects reading times. She creates sentences that are neutral, positive, or negative; she controls for word frequency, sentence length, and syntactic structure. Why does she control for word frequency, sentence length, and syntactic structure of sentences?
Because these variables may confound the effect of emotional valence on reading times
Using the partial Latin square counterbalancing method, I generate the following 3 orders for 4 conditions (A, B, C, D) of an IV: 1. ABCD 2. BCDA 3. CDAB What's the 4th order?
DABC
In a design with 2 IVs, 1IV has no main effect and IV2 has no main effect. I can therefore conclude that there is no interaction of the two factos
False
T/F? In between-subjects design, a given participant is assigned to all conditions/levels of an independent variable
False
T/F? Internal validity applies to observational studies
False
what is the best order to present the descriptive and inferential statistics
It depends on you research design and your main hypothesis of interest
In the scenario discussed, is there a main effect of intensity of exercise?
No
What's one disadvantage of between-subjects designs?
Participants may have individual differences that may be relative to the study and thus are a confound.
Which of the following is NOT an order effect that may be observed in a within-subjects design with two conditions?
Performance declines across conditions because there is a real difference between conditions
A researcher is interested in examining whether the emotional valence of sentences affects reading times. She wants to use a within-subjects design. Since the order of valence conditions (positive, negative, neutral) can matter in a within-subjects design, what's a solution?
Researchers can counterbalance so different participants experience the three conditions in different orders.
You the following statistics in a paper reporting a correlational study. What should you conclude? r (28)=-.33, p=.18
There is not a significant correlation
Reliability refers to the extent to which a given measure produces similar results under the same conditions
True
T/F? One of the benefits of factorial designs is that they permit examining how the effect of one factor depends on another factor
True
Using minor deception is one way of controlling for participant biases
True
When conditions differ due to selection bias, these systematic differences across conditions are a confound that compromises validity
True
A researcher is interested in examining whether the emotional valence of sentences affects reading times. Since the order of valence conditions (positive, negative, neutral) can matter in a within-subjects design, what's a solution?
Use a between-subjects design: each participant reads sentences of only one type (positive, negative, or neutral)
One research methods section got a mean of 16.60 out of 20 on their quiz #1; another section taught by the same instructor got a mean of 15.30 on the same test. What can we conclude?
We don't have enough information to conclude anything
In this scenario, is there an interaction between intensity and duration of exercise?
Yes
In a two-group between-subjects design with matched groups, we would use:
a paired samples T-test
What is true in *balanced* Latin square counterbalancing design?
a. Each condition appears only once in each position of the sequence of conditions (i.e., in 1st, 2nd, 3rd, 4th position) b. Each condition precedes and follows each of the other conditions an equal number of times
A researcher is interested in examining whether the emotional valence of sentences affects reading times. She creates sentences that are neutral, positive, or negative (she controls for word frequency, sentence length, and syntactic structure). a. In a between-subjects design, how would the conditions of emotional valence be distributed across subjects? (i.e., what would a given subject experience?) b. In a within-subjects design, how would the conditions of emotional valence be distributed? (i.e., what would a given subject experience?)
a. For a between-subjects design, each participant would experience only one level (positive, neutral, or negative) of the independent variable (emotional valence of sentences). b. For a within-subjects design, each participant would experience all levels (positive, neutral, or negative) of the independent variable (emotional valence of sentences).
Imagine that you are studying the relationship between the presence of violence in television programs and aggressive behavior exhibited in a computer-based task. Suppose that the violent television scenes you found are all louder than those with nonviolent scenes. a. What is a potential confound here?b. How can you control it?
a. the volume of the violent scenes may be a confounding variable b. This can be controlled by lowering the volume of the violent scenes to match the non-violent scenes.
What is NOT a way of minimizing experimenter biases?
asking experimenters to motivate participants in one condition more than participants in other conditions
The extent to which a variable accurately reflects or measures the behavior of interest is called
construct validity
The aspects of a research study that help participants work out what is expected of them, and consequently lead them to behave in unnatural ways are called:
demand characteristics
Researchers have expectations about the participants' performance in the study and may unconsciously influence participants. What is one way of controlling or minimizing these effects?
double blind
When researchers unconsciously behave in a way that influences participants' behavior, this is called
experimenter biase
What should you NOT do in your results section
explain why you're using the inferential test you are using
If the findings of a study on how people cooperate in the lab are consistent with those of a study examining how people cooperate in an organization, this means that the original study had high
external validity
Making claims about human psychology when only using samples of undergraduates in the US poses a threat to
external validity
The extent to which on can generalize from the research setting and participant population to other settings and populations is called
external validity
The extent to which one can generalize from the research setting and participant population to other settings and populations is called
external validity
If an experiment has more than one independent variable, the design is:
factorial design
when studying the cognitive process of language use, which approach has more external (ecological) validity
having participants discuss some stimuli without contraints on what they can say
What is NOT true about between-subject designs
independent variables in between-subjects designs can only have 2 conditions
A t-test is a
inferential statistic
In experiments, which aspect of validity do confounds compromise the most?
internal validity
Which type of validity permits attributing a causal relationship between the IV and DV?
internal validity
What is NOT a benefit of matching prior to random assignment?
it increases Type 1 error
In a between-subject design, we can make participants in different conditions more comparable by
matching them on the basis of some potentially confounding variable
What's one disadvantage of within-subjects designs?
order effect
What are the 3 subsections most typically found in the METHOD?
participants, materials (or measures), and procedure
When each participant has an equal chance of being assigned to different conditions the selection process is called
random
The extent to which a measure produces similar results under the same condition is called
reliability
If a student participant completes a given measure at two different time points and scores similarly, that measure has good
test-retest reliability
In the scenario described, and interaction of the intensity and duration of yoga exercise means that
the effect of exercise duration on the reduction of depression symptoms depends on the level of intensity
Construct validity refers to
the extent to which a test measures the construct of interest
In a version of the described study in which each participant is asked about the subjective distance from Bay station to each of the 4 stations of interest
the independent variable station identity would vary within participants
A group of researchers study whether the frequency of children's pointing gestures predicts their language development. Part of the study involves identifying (coding) children's pointing gestures in videotaped sessions with a caregiver. By having two research assistants code a subset of the same videos for the presence of pointing gestures, the researchers establish:
the interrater reliability of the coding scheme used to operationally define pointing gestures
What kind of information should NOT go into a table
the raw data
You see a print out with the informtation below for a two-group experiment: t(24) = 3.37, p<.01 what should you conclude?
there is a significant difference between your 2 means
Which is NOT true about factorial designs?
they involve one independent variable with two or more levels (conditions)
What is NOT true about establishing interrater reliability
two or more raters must code (or rate) a subset of the same observations while consulting with on another
A researcher is interested in examining whether the emotional valence of sentences affects reading times. She creates sentences that are neutral, positive, or negative (she controls for word frequency, sentence length, and syntactic structure). How should the researcher analyze her data to examine the effect of emotional valence on reading times?
with a one-way ANOVA (Analysis of Variance)
