ap psychology: research unit

0.0(0)
studied byStudied by 3 people
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
Card Sorting

1/83

Study Analytics
Name
Mastery
Learn
Test
Matching
Spaced

No study sessions yet.

84 Terms

1
New cards

theory

possible explanation of a phenomenon → helps validate

example - the more you see someone, the more you like them

THEORY AND HYPOTHESIS ARE NOT THE SAME

2
New cards

hypothesis

testable prediction of an outcome (or a study) → helps form

example - subjects will more likely want to play a game with people they saw earlier opposed to people they just met

3
New cards

operationalization(operational definition)

how is it measured?

(is the measure good/fair/makes sense?)

how variables in a study are measured

4
New cards

types of measures

survey, test, inventory

5
New cards

validity

is it accurate?

6
New cards

reliability

is it CONSISTENT?

7
New cards

population

group of people who possess a certain characteristic (who I may want to study)

8
New cards

sample

smaller group that represents the population

the group participating in the study

9
New cards

internal validity

measure assesses what it is supposed to

10
New cards

external validity

sample represents the population → results from sample can be applied to population

11
New cards

random sampling

people from the population who are chosen randomly → good external validity

12
New cards

sample bias

poses the greatest threat to external validity → poor external validity

13
New cards

case study

  • poor external validity

  • rare concept with much info

  • provides the most detailed qualitative information

  • allows you to collect a lot of information from something rare or unusual that is likely not possible with other research methods

  • a type of a research method that you can’t really conduct repeat and experiments on so you use the all information that is already available

14
New cards

meta analysis

uses previous studies and puts it all together

15
New cards

experiment

  • study 1, study 2

  • variables that can be manipulated

16
New cards

observational study

  • naturalistic → variables you can’t control

  • laboratory → artificial

17
New cards

pro and con of experiment

infer cause and effect

can be complex

18
New cards

variable: dependent

  • it’s affected/depends on the independent variable

  • you can’t control

19
New cards

variable: independent

the variable you can control and change

20
New cards

variable: control/controlled

the variable you keep the same throughout the experiment

21
New cards

variable: extraneous

variables you can’t control

22
New cards

confounding

  • interferes with the influence of the independent variable

  • a kind of extraneous variable that is messing things up; makes things confusing

  • variables that adversely affect an outcome of a study that we cannot control and/or are not aware of

23
New cards

the effect of extraneous/confounding variables

researchers take into account extraneous variables because the whole experiment must be considered invalid if another variable affects the study because the experiment is longer testing one’s variable’s affect on the other

24
New cards

control group

no special treatment

25
New cards

experimental group

gets special treatment

26
New cards

ethics

  • IRB approval

  • confidentiality

  • informed consent

  • no coercion

  • minimized risks

  • deceptions when necessary

  • debriefing

27
New cards

IRB (Institutional Review Board)

  • approves any kind of research before it starts

  • reviews research purpose

  • will accept or reject and ask you to revise

  • looks at research ethics

  • they do NOT fund the study

28
New cards

confidentiality

the participant can tell whoever but the research has to make sure the participant’s identity is kept a secret unless it’s an emergency like they’re going to hurt themselves

29
New cards

informed consent

participant has to give you their consent to do the experiment unless you’re in a public space like a mall

30
New cards

no coercion

researcher cannot force a participant to do anything nor can they offer money

31
New cards

minimized risks

ALL research has risks but it is always better to have less risks

the IRB asks “does the benefits outweigh the risks?” example - coronavirus vaccine

32
New cards

deception when necessary

it is permitted under certain circumstances

33
New cards

debriefing

tell participants about all the aspects of the study (what is going on, the purpose, hypothesis, etc.) following participation in a study

34
New cards

random assignment

each person has an equal chance of being in either the control or experimental group

avoids experimental bias

35
New cards

double blind procedure

neither the researcher nor the participant knows the participant assignment

avoids experimental bias

36
New cards

single blind procedure

the researcher knows the assignment of the participant while the participant does not know their assignment

37
New cards

experimenter bias

  • researcher influences results

  • sometimes deliberate, sometimes not

38
New cards

social desirability bias

appealing better than one is in a study

examples - rating oneself better, knowing like you have knowledge

39
New cards

placebo effect

“fake pill” → healing response

expectation of healing → healing response

40
New cards

hawthorne effect

knowing you’re being studied which facilitates behavior

41
New cards

correlation

the relationship between two variables

NOT CAUSATION (just because there’s a correlation between two variables does not mean one variable caused the other variable to happen)

42
New cards

spurious correlation

  • a misleading/non-existent relationship

  • two variables that appear to correlate but in fact do not

43
New cards

positive correlation

same direction: ABand A B

POSITIVE AND NEGATIVE DO NOT REFER TO GOOD AND BAD

44
New cards

negative correlation

opposite direction: ABand A B(inverse)

POSITIVE AND NEGATIVE DO NOT REFER TO GOOD AND BAD

45
New cards

correlation coefficient

  • direction → positive (+) or negative (-)

  • strength of relationship → how much do the dots look like a line? (number from 0 to 1)

46
New cards
<p>0.2</p>

0.2

weak correlation, dots are going in a positive up direction

47
New cards
<p>-1.0</p>

-1.0

perfect correlation, dots made a straight line that is going in down negative direction

48
New cards

mean

  • the average

  • most easily affected by extreme scores

49
New cards

median

  • middle value

  • is the middle of all vales

50
New cards

mode

  • most frequent mode

  • can be more than one

  • mode is the highest point

51
New cards
<p>normal distribution</p>

normal distribution

mean=median=mode

52
New cards
<p>positive skew</p>

positive skew

mean>median>mode

53
New cards
<p>negative skew</p>

negative skew

mode>median>mean

54
New cards

measures of spread

=measures of variation

55
New cards

range equals

highest-lowest

56
New cards

standard deviation

  • standard measure or variability

  • can compare one variable to another

57
New cards

standard deviation equals

square root of variance

the higher the SD, the bigger the spread

58
New cards

variance equals

square of standard deviation

☆ the higher the SD, the bigger the spread

59
New cards

EX: 100 mean (0 SD)

1 above: 110

1 below: 90

SD is 10

60
New cards

z-score

number of standard deviations above or below the mean

61
New cards

z score equation

z score = score-mean/sd

62
New cards

if the mean of test scores was 100 and the standard deviation was 15, what is the z-score of a person who got 130?

130-100=30 30/15=2

63
New cards

(consider a normal distribution) given a mean of 50 and a standard deviation of 5, what percentage of participants had a score between 45 and 55?

68%

64
New cards

percentile

what percentage of population did a value outscore? → add up everything to the left of the value

65
New cards

percentage is the same as _____

percentile

66
New cards

suppose a person has a score of 1.0. what is that person’s percentile?

84

67
New cards

what is the percentile of someone who score corresponds to a z-score if -1?

16th percentile (or 15.5 if you don’t round)

68
New cards

concluding an effect or difference exists →

conclusion is wrong →

type 1 error → chance of this → p-value

69
New cards

any p-value equal or less than this is acceptable

0.05

70
New cards

statistical significance

→ low type 1 error probability

→ confidence that difference/effect exists

→ “not due to chance”

71
New cards

t-test

compares means of two groups

72
New cards

ANOV

compares means of 3 or more groups

73
New cards

hindsight bias

“I knew it all along” ← after conducting research and coming to a conclusion, you feel like this

74
New cards

confirmation bias

belief → seeks evidence that supports

belief → ignores evidence that goes against

75
New cards

construct validity

theoretical concept

does the test accurately assess the theoretical concept?

76
New cards

criterion validity

specific skills

does the test accurately measure a certain set of skills?

77
New cards

content validity

measure that accesses how well a test or other assessment tool covers all relevant aspects of a construct

do the test items collectively fully cover the theme being assessed?

78
New cards

concurrent validity

measure that accesses how well 2 tests or assessments agree when they were taken at the same time

does the test correlate well with an already established test?

79
New cards

predictive validity

ability of a test or measurement to predict a future outcome, such as behavior, performance, or disease

80
New cards

face validity

does the test look like a legitimate test?

81
New cards

test- retest reliability

does administering the SAME test give you similar results?

82
New cards

parallel forms reliability

do different test versions with similar content yield similar results?

83
New cards

inter-rater reliability

do obsevers of the same event rate it the same way?

84
New cards

split half reliability

is there consistency WITHIN the test?