EXAM 1 NOTES Flashcards

Question

What validities are important for association claims?

Answer 1

Statistical validity (correlation coefficient: tells us how related they are, the strength and direction of the relationship) Construct validity External validity Internal validity is NOT important for association claims because it’s not saying one causes the other so we dont need to rule out other confounds

Answer 2

Internal validity is most important Construct validity Statistical validity External validity is least important and will often be sacrificed to increase internal validity

Answer 3

Causal claims are the STRONGEST claims to make. Criteria for causal claims Covariance: x and y have to be associated Internal validity: rule out confounds Temporal precedence: Prove that x caused y because if y happened first, theres no way x caused y

Answer 4

Yes because the variables need to measure what you intend to measure

Answer 5

Respect for person’s: Informed consent. In this study, participants were lied to; they were told they would be treated for their condition and they weren't even told they had syphilis; they were told they had bad blood Beneficence: maximize benefits and minimize harm. In this study, they did not give participants treatment. Instead, the study wanted to see how syphilis would progress without treatment. This causes many to get very sick, die, and spread the disease to their loved ones. Justice: Who is in the study, and who is benefiting? In this study, participants were only poor black men, but anyone can suffer from syphilis. They suffered to benefit the rest of the population

Answer 6

Respect for persons: Lied to participants by saying someone was being shocked, and they weren't. Beneficence: Psychological effects on participants guilt and anxiety for thinking they were causing harm to someone. Justice: no violation

Answer 7

Reports on original research they are primary sources

Answer 8

Google scholar and PycInfo

Answer 9

Abstract: summary of the article Introduction: provides background information Method: details on how the study was conducted Results: presents the findings of the study Discussion: interprets the results and their implications References: what other sources were used to write the article?

Answer 10

Currency: how recently was it published Relevance: who is the intended audience? Does it relate to what you are looking for? Authority: Who is providing the information? What are their credentials? Are they an expert? Accuracy: Is the information supported by evidence? Can you verify it through other sources? Purpose: What purpose does the information serve? Is it to inform, persuade, or market?

Answer 11

Self-report: people provide their own information about their thoughts, feelings, behaviors, and experiences example: asking someone how many anxious moments they’ve had this week Observational: collecting information through observation example: fingernail biting or stuttering during a conversation Physiological: measuring bodily functions. Example: measuring cortisol levels in saliva

Answer 12

Categorical / Nominal: qualitative, levels are distinct, order doesn’t matter (example: gender: male, female, nonbinary) Ordinal: ranked order, distance between levels doesnt matter (example: 1st, 2nd, 3rd, the time difference between 1st and second could be 5 seconds and the time distance between 2nd and 3rd is 8 seconds) Interval: equal distance between levels, no true 0 (example: jeans sizes: 0,2,4,6,8 etc. A size 0 doesnt mean the person has no waist.) Ratio: equal distance between levels, 0 is meaningful. (example: asking someone how many cats they have. If they have 0, they have none, if they have 3 they have 3)

Answer 13

All variables must have two levels of measurement but the levels can differ

Answer 14

Test retest: consistent scores every time the measure is used (example: someone will have about the same score today, tomorrow, next week, or next month) Interrater: consistent scores no matter who measures. (example: rater 1 get’s a very similar score to rater 2) Internal: Regardless of how the question is phrased, the participant will give the same answer

Answer 15

Face: It looks like what you want to measure (example: If a math test looks like a math test, then it has high face validity) Content: It contains all the parts that your theory says it should contain (example: the math test only contains addition and no subtraction, multiplication, or division). It therefore has low content validity. Criterion: Demonstrate that your measure can predict the key outcomes. Using correlations, or known groups: find groups that you know will be different and measure your variable in those groups (e.g., depressed and not depressed); measure BDI and it should be higher in depression group Convergent criterion validity: High correlations between two tests that measure the same or similar things Divergent criterion validity: Low correlations between a test and something it is not supposed to measure

Answer 16

No, some things can be reliable and not valid but things CAN NOT be more valid than it are reliable Example: head circumference as a measure of intelligence is very reliable; everytime they’ll have the same size head but it won’t measure intelligence

Answer 17

Construct validity is more importantly demonstrated in general constructs

Answer 18

Both of these are seen during observational measures Observer bias: when the participant isn’t acting different but based on the researcher’s bias they’ll record the participants behavior as different Observer effects: when the observer effects the behavior of the observed

Answer 19

Yes the beginning questions can impact future responses in a biased way

Answer 20

people give their opinion by picking the best of two or more questions (used in political polls)

Answer 21

strongly agree, agree, neither, disagree, strongly disagree

Answer 22

a numeric score paired with adjectives ex: rate my professor level of difficulty 1: show up and pass 5: hardest class ever

Answer 23

yes use simple questions because people get confused and it may lead to different results

Answer 24

For each question, word it in different ways and if you get the same results, the question is worded fine

Answer 25

Asks 2 questions in 1; these have poor construct validity. people may only answer one part of the question instead of both so its best to separate it

Answer 26

Instead of getting info on their opinion, it tests their working memory. whenever a question contains negative phrasing, it can cause confusion

Answer 27

Self-reports can be ideal but can also be imperfect due to response sets (getting lazy, especially during a long questionnaire and clicking all neutral or all positive or negative) response sets weaken construct validity

Answer 28

Acquiescence, aka yea-saying: saying yes or strongly agreeing without thinking carefully fence sitting: playing it safe by answering the middle of the scale (weaken construct validity) to combat this, sometimes researchers will take away the neutral option (the downside to this is that sometimes people wont answer without a neutral option)

Answer 29

including reverse items: changing the wording to mean the opposite of a previous question (this also may slow people down and make them answer more carefully because they recognize the question) the reverse-worded items have higher construct validity

Answer 30

Answering differently than how you feel to make yourself look better. To avoid this, the researcher may make it anonymous (this may also cause them to take it less seriously if its anonymous), also including statements that target socially desirable responders ex: my table manners at home are as good as when i eat at a restaurant

Answer 31

For the most part, its accurate, but for certain circumstances, its not. Ex: traumatic events, or years ago at times (9/11 people recall watching the first plane hit on TV, but it was not on TV yet).

Answer 32

Using multiple observers, although this does not ensure there won't be bias, as well as masked or blind design where the observers are unaware of the purpose of the study and the conditions to which participants have been assigned.

Answer 33

people change their behavior when they know theyre being watched, which also occurs in animal subjects

Answer 34

make yourself less noticeable (as the observer) this is to avoid observer bias

Answer 35

yes but if you record them at the end of the study, you need to delete the footage if they arent okay with it

Answer 36

knowing what kind of claim is being made, knowing whether to believe it, knowing how to apply it, and knowing when to invest time money and effort into a program (e.g., the Scared Straight program seems effective, but its actually harmful)

Answer 37

”Venting anger through physical means releases the anger from your body”

Answer 38

the manipulated variable is controlled the measured is observed and recorded

Answer 39

some can be both or one or the other; ex: height and age are measured; you can't manipulate that, Medication dose or number of steps walked can be measured or manipulated

Answer 40

the value that is the result of your analysis

Answer 41

margin of error of the estimate. a range where the true value lies.

Answer 42

Do other studies yield the same or similar results?

Answer 43

it is justified but only to the extent that it is necessary to achieve the goals of the study

Answer 44

omission: withholding details from participants commission: lying to participants

Answer 45

when deception is used, researchers must debrief participants at the end of the study: explain why deception was used and the nature of the deception

Answer 46

replacement: find alternatives to animals in research when possible refinement: modify procedures to minimize distress reduction: use designs that require the fewest animal subjects as possible

Answer 47

scatterplots and statistically evaluate reliability using correlation coefficients

Answer 48

average of all correlations between different items

Answer 49

interrater reliability and test-retest reliability