Psychology - RESEARCH METHODS

?
distribution curves - NORMAL
symmetry at the middle, end points meet x axis, bell shaped, measure of central tendancy are the same, this is expected from normal target population
1 of 88
distribution curves - POSITIVE SKEW
more low scores than high, measures of central tendancy decrease
2 of 88
distribution curves - NEGATIVE SKEW
more high scores than low, measures of central tendancy increase
3 of 88
Errors - 2 types and why
type 1 = false positive, type 2 = false negative; happen when significance levels are set too low or high
4 of 88
significance levels
determine whether we accept or reject the hypotheses; most common is p
5 of 88
probability =
desired outcome/total possible outcomes
6 of 88
symbols
< less than, > greater than, , ~ approximately, ∝ proportional
7 of 88
my practicals - experiment
controlled setting, chewing gum affect concentration, opportunistic, independant measure, had to use deception, anonymous and debriefed afterwards, may have been individual differences - practice, intelligence
8 of 88
my practicals - self report
do colours affect mood/emotions; open and closed 8 questions, opportunistic, anonymous and written consent, didn't always answer questions correctly and i didn't find this in my pilot study
9 of 88
my practicals - correlation
gcse score and age; opportunistic, weak positive correlation, wasn't significant because didn't meet calculated critical value for n=15
10 of 88
my practicals - observation
observed flirting behaviour, non pp + covert (to maintain e.v and prevent dc), structured - coding frame, event sampling, inter rater was 0.96, could improve this by explaining behaviours clearly and training observers beforehand with pilot study
11 of 88
types of observations - 8
naturalistic = natural setting of pps; controlled = control all factors affect behaviour; structued = coding frame; unstructured = record everything; cover = undercover; overt = know and consented; participant or non pp
12 of 88
designing observations - 3 types and whats used
time point sampling = record whats happening at certain intervals; time event sampling = record everything within a set time; event sampling = record behaviours each time they happen; we use coding frames
13 of 88
self report - 4 types (first 2)
questionnaire - gathers lots of data quickly, may be cultural limitations, low response rate; structured interview - data easy to compare, standardised, pps can't explain/elaborate;
14 of 88
self report - 4 types (second 2)
semistructured - clarify/explain, both types of data, not standardised; unstructured - can build rapport, rich data with explanations, time consuming so may use less pps?, difficult to analyse, higher ecological validity
15 of 88
experiments - what are their features ?
have IV and DV, hypothesis which predicts, aims to establish cause and effect
16 of 88
experiments - lab
controlled setting; eliminates EV's, can test replicability so is reliable, lacks ecological validity
17 of 88
experiments - field
natural environment of pps; higher ecological validity, less likely to alter behaviour, can't control EV's and isn't standardised
18 of 88
correlations
can be positive or negative; look for relationship between 2 variables; allow us to investigate things couldn't do ethically in experiments, but correlation doesn't mean cause and effect
19 of 88
experiments - quasi
IV is naturally occurring and can't be directly manipulated by researcher; allows us to measure how these IVs affect behaviour, little sample bias, but may be variance in pps due to different upbringings, hard to replicate
20 of 88
Extraneous variables
can be participant (e.g. intelligence) or situational (e.g. room temp); a confounding variable is one that hasn't been eliminated or controlled and so affects the DV
21 of 88
hypotheses - what are they and 3 types
arise from previous research or an event; can be alternate: one tailed (directional) or two tailed (non directional), or null
22 of 88
how to get a sample
general population -- target population -- sample that is representative of who we're going to generalise to
23 of 88
types of sampling - opportunity
use whoevers available; quick and easy, not very representative
24 of 88
types of sampling - self select
volunteers from adverts; know they will fit the criteria, but may have strong opinions = bias
25 of 88
types of sampling - random
everyone in TP has equal chance of being chosen; very representative but may give unequal gender distribution
26 of 88
types of sampling - snowball
one person from odd group is chosen and they suggest others; more ethical than asking prying questions, but may not reply and are all from same area as first person
27 of 88
experimental designs - independant measures
randomly allocate pps to IV; simple and quick, no order effects/demand characteristics, may be individual differences
28 of 88
experimental designs - repeated measures
tested on all IV levels; more data, may be order effects/demand characteristics, no individual differences
29 of 88
experimental designs - matched pairs
pps matched on similar factors then split; trys to eliminate individual differences, no order effects/demand characteristics, time consuming and may be hard to match exactly
30 of 88
demand characteristics
pps act unnaturally as they know they're being researched; can be a confounding variable
31 of 88
social desirability - what can we use to check for it
pps don't answer truthfully, but answer to appear a better person; can use face validity to check for this
32 of 88
researcher effects
things researcher does that affects quality of data
33 of 88
researcher bias
attitudes/actions that threaten quality of research as a whole
34 of 88
3 types of self reports
rating scale - measure variables e.g 1-10; likert scale - choose response that matches feeling, e.g agree/disagree; semantic differential scale - choose point between 2 opposites, e.g expensive----cheap
35 of 88
sections of a report - the order
abstract, intro, method, results, discussion, appendices, references
36 of 88
sections of a report - abstract
brief summary and what was found
37 of 88
sections of a report - intro
area of psychology, aims and hypotheses
38 of 88
sections of a report - method
sample, design, procedure, apparatus
39 of 88
sections of a report - results
raw data and summaries of it
40 of 88
sections of a report - discussion
evaluation of results, alternate explanations
41 of 88
sections of a report - appendices
calculations/tables
42 of 88
sections of a report - references
use harvard referencing system - name of researcher, date of publication, title of study, title of journal, volume, page numbers
43 of 88
peer review
work sent to others with expertise in the field; can either approve for publication or give recommendations for improvement, maintains quality of journals, prevents plagiarism/fabrication, may be biased - might not approve if don't agree
44 of 88
how science works - cause and effect
research can show one factor causes another; improve with controls/standardisation
45 of 88
how science works - falsification
testing other hypotheses and proving them wrong makes our study more reliable
46 of 88
how science works - replicability
study can be repeated and get the same results; improve with controls/standardisation
47 of 88
how science works - hypothesis testing
test using observation/experiments to gain new knowledge
48 of 88
how science works - objectivity
researcher isn't influenced by personal opinions, no bias
49 of 88
how science works - manipulation of variables
see if results change with/without presence of IV
50 of 88
how science works - quantifiable measures
quantitative, objective, observable, scientific
51 of 88
how science works - controls
eliminate evs to be sure cause and effect is true
52 of 88
how science works - standardisation
control apparatus, procedure, instructions to control situational evs - makes it replicable
53 of 88
how science works - induction
bottom up theory; think of hypotheses and test them, look for patterns then develop theory
54 of 88
how science works - deduction
top down theory; think of theory, create testable hypotheses then test them to prove our theory
55 of 88
levels of measurement - nominal
NUMBER of pps in categories
56 of 88
levels of measurement - ordinal
data can be ORDERED from highest to lowest
57 of 88
levels of measurement - interval
data measured in equal intervals
58 of 88
levels of measurement - ratio
interval, but with a true zero - time
59 of 88
reliability - define
where a study can be used more than once and produce consistent results
60 of 88
reliability - internal
within, first half same as second? split half test
61 of 88
reliability - external
gives same results over time; test retest
62 of 88
reliability - inter rater
2 observers get same results, shows there's no observer bias
63 of 88
validity - define
how accurate research is at measuring what it aims to
64 of 88
validity - internal
ev's aren't causing an effect
65 of 88
validity - external
can be generalised due to being conducted realistically
66 of 88
validity - face
how good it appears to be
67 of 88
validity - construct
measure actual behaviour it aims to
68 of 88
validity - concurrent
gives same results as other tests measuring the same thing
69 of 88
validity - criterion
how accurately one measure predicts another
70 of 88
validity - population
how accurately measures behaviour in general population
71 of 88
validity - ecological
same as real life
72 of 88
measures of central tendancy - mean
uses all data values so is most sensitive; don't use with extreme anomalies
73 of 88
measures of central tendancy - median
is less sensitive to variations; less sensitive so may not show true picture
74 of 88
measures of central tendancy - mode
good for data in categories; may be no/more than one mode
75 of 88
types of data - primary
conducted by researcher, find exactly what we want, may not be as accurate as large companies with access to lots of pps
76 of 88
types of data - secondary
published/official statistics, accurate and reliable, may be dated
77 of 88
charts - bonus
describes data and makes it easier to analyse
78 of 88
charts - pie
shows data as a percentage of a whole
79 of 88
charts - line
shows 2 conditions over a period of time
80 of 88
charts - bar
data from table; not continuous
81 of 88
charts - histogram
data from table; continuous
82 of 88
charts - scatter
shows relationship between 2 variables
83 of 88
ethical considerations - what and why?
CRIR; ethics are the science of morals; give psychology a good name and means pps with take part again
84 of 88
ethical considerations - C
competence - understand limits of knowledge, accept responsibility to try and resolve issues
85 of 88
ethical considerations - R
respect - respect cultural differences, right to withdraw, confidentiality, informed consent
86 of 88
ethical considerations - I
integrity - value honesty, avoid deception
87 of 88
ethical considerations - R
responsibility - ensure pps come to no physical/psychological harm; used debriefing if they do
88 of 88

Other cards in this set

Card 2

Front

distribution curves - POSITIVE SKEW

Back

more low scores than high, measures of central tendancy decrease

Card 3

Front

distribution curves - NEGATIVE SKEW

Back

Preview of the front of card 3

Card 4

Front

Errors - 2 types and why

Back

Preview of the front of card 4

Card 5

Front

significance levels

Back

Preview of the front of card 5
View more cards

Comments

No comments have yet been made

Similar Psychology resources:

See all Psychology resources »See all Research methods and techniques resources »