Research Methods year 1 Flashcards
(85 cards)
1
Q
Lab Experiment
A
Experiment conducted in a well controlled setting where the researcher manipulates IV and records effects on DV
2
Q
Strengths + Limitations of lab experiment
A
- Easy to replicate which makes it more reliable (S)
- Extraneous variables can be controlled (S)
- May lack mundane realism or ecological validity so it cannot be generalised (L)
- Demand characteristics may be present and bias the results (L)
3
Q
Field Experiment
A
Conducted in a uncontrolled setting where the IV is manipulated and effects on DV are measured
4
Q
Strengths + Limitations of field experiment
A
- More ecologically valid (S)
- Higher mundane realism (S)
- May reduce demand characteristics if covert (S)
- Difficult to replicate so less reliable (L)
- Extraneous variables are harder to control=lack of validity(L)
5
Q
Natural Experiment
A
Experiment where the IV cannot be manipulated by the researcher as it happens anyway and the DV is an impact of the IV
6
Q
Strengths + Limitations of natural experiment
A
- Allows study of things which usually wouldn’t be able to be studied due to ethical and practical reasons (S)
- Highest ecological validity as it is a real life situation(S)
- May be rare conditions and cant be generalised (L)
- No control of extraneous variables so IV may not impact DV
7
Q
Quasi-Experiment
A
Experiment where the IV is natural and re-occuring (eg age or gender) and impacts the DV
8
Q
Strengths + Limitations of quasi-experiment
A
- Often conducted in lab conditions so high replication and reliability (S)
- Control of extraneous variables (S)
- Random allocation cannot be used so confounding variables are more likely to be present (L)
9
Q
Experimental Designs
A
Matched Pairs
Repeated Measures
Independent Groups
10
Q
Matched Pairs Design
A
Different participants are used in each condition but are matched in pairs depending on common characteristics such as age then randomly allocated
11
Q
Strengths + Limitations of matched pairs
A
- Avoids order effects (S)
- Minimises participant variables (S)
- Time consuming as its hard to find a match + more money (L)
- More materials required (L)
12
Q
Independent Groups Design
A
Different people who are randomly allocated to different conditions
13
Q
Strengths + Limitations of independent groups
A
- Avoids order effects(S)
- Less likely to guess the aim and give demand chracteristics (S)
- Participant variables as lots are needed (L)
- Time consuming + more money (L)
14
Q
Repeated Measures Design
A
Same participants used in both conditions
15
Q
Strengths + Limitations of repeated measures
A
- Money + time efficient as less participants are required(S)
- Minimises participant variables as the same ones are used (S)
- Order effects may be present (L)
- Demand characteristics may bias results (L)
16
Q
Sampling Methods
A
Random sampling Systematic sampling Stratified sampling Volunteer sampling Opportunity sampling
17
Q
Random sampling
A
Participants are randomly chosen so the population has an equal chance to picked eg picking names out of a hat
18
Q
S + L of random sampling
A
- Bias is unlikely as researcher has no control (S)
- Can be time consuming and difficult to obtain good target pop(L)
- Chosen ppt may refuse to partake
- May still be biased and unrepresentative by chance (L)
19
Q
Opportunity Sampling
A
Researcher selects participants depending on who is available at that time
20
Q
S + L of opportunity sampling
A
- Money + time efficient as its quick and easy (S)
- Often less generalisable as its only a small group from the population (L)
- May be subject to researcher bias as they choose (L)
21
Q
Volunteer Sampling
A
Participants put themselves forward to be put into the sample
22
Q
S + L of volunteer sampling
A
- Money + time efficient as participants come to the research, less effort needed (S)
- May only attract a certain profile or type of person (eg a helpful person) so volunteer bias makes it ungeneralisable.
23
Q
Stratified Sampling
A
Is a proportional representation of the target population as it is broken down into strata (sub groups) like age or gender
24
Q
S+L of stratified sampling
A
- Can be very representative and generalisable as the whole population would be equally represented (S)
- Avoids researcher bias as random allocation occurs within strata (S)
- Strata may not represent all different groups in target population (L)
- Time consuming + more money required (L)
25
Random Allocation
Participants are randomly assigned to different conditions to avoid participant variables (often in independent groups)
26
Counterbalancing
Half the participants do the conditions in one order than the other half do the opposite order (eg: 1/2 do A-B, 1/2 do B-A) there will still be order eff, results will just be more representative and general as an average
-Controls for order effects in repeated measures
27
Demand Characteristics
Participants guess the aim and act accordingly, please you or screw you effect, either way the results wont be valid
28
Extraneous Variables
Variables other than the IV which have the potential to effect the DV if not controlled (eg temperature, environment)
29
Confounding Variables
Variables related to IV which do effect the DV so the researcher cant be sure whether the IV effected the DV or whether it was just the confounding
30
Investigator effects
Any effect from the researchers behaviour whether its conscious or unconscious
31
Overt experiment
Participant knows they are being observed
32
Covert experiment
Participant doesn't know they are being observed
33
Naturalistic Observation
Takes place in a place where the behaviour would naturally occur
34
Controlled Observation
Takes place in artificial and controlled setting so extraneous variables can be controlled
35
Single blind design
Where the researcher knows fully what they are doing but the participant doesn't
36
Double blind design
Where the study is carried out by an independent researcher and the first researcher and participant doesn't know, this avoids researcher bias
37
Standardisation
Using exactly the same instructions and procedures for all participants
38
Randomisation
Use of chance in order to control for effects of bias when deciding materials and order of conditions
39
Pilot Study
Researcher carries this out before the actual research to adjust and fix any flaws
40
Overcoming informed consent
- Presumptive consent: asking similar group of people and seeing if they consent then assuming actual participant consent
- Prior general consent:participants consent to several studies including one with deception, if they consent, they consent to be decieved
- Retrospective consent: Asked for consent during debriefing, they may have already taken part without knowing
41
Overcoming deception (when info is withheld from participants or are misled)
Debrief them: Reveal everything about the study, aims, procedure, their role and how their data will be used at the end
42
Overcoming physical and psychological harm
Right to withhold data- not let researcher use their data
| Counselling- Therapy to revive mental state to the same as prior to the study
43
Overcoming confidentiality
-Keep participants anonymous eg refer to them as numbers or use initials
44
Participant observation
The researcher takes part in activities and involves themself whilst also observing behaviours
45
Non-Participant observation
Researcher does not take part and involve themself, they just observe
46
Self-report techniques
Method of gathering data where participants provide info about themselves
- Questionnaires
- Unstructured and structured interviews
47
Unstructured interview and S + L
A conversation with no set questions, often free flowing and open ended questions
- Questions can be followed up and more detail (S)
- Understand ppt perspective better + less misinterpretation of questions (S)
- Data often qualitative and harder to analyse (L)
- Time consuming (L)
48
Structured interview and S+L
Fixed and set questions often closed questions.
- Easier to replicate as its like a questionnaire (S)
- Less subjective than unstructured int as everyone answers the same questions (S)
- Lack of detail and elaboration (L)
- More room for misinterpretation (L)
49
Open and closed questions S+L
Open=Qualitative data, more depth(S), demand characteristics may occur (L)
Closed=Quantitative data, Easy to analyse(S), time efficient(S), lacks depth(L)
50
Types of correlation
Positive- When both variables increase or decrease at the same time
Negative- Where one variable increases and the other decreases
Zero-No relationship between co-variables
51
Difference between experiment and correlation
An experiment manipulates the IV to see if it has an effect on the DV
Correlation looks for the type of relationship between the IV and DV
52
Correlation Coefficients
-1 (negative corr) to +1 (positive corr)
0-0.3 weak
0.3-0.7 moderate
0.7+ strong
53
Case studies and S+L
In depth study about a individual, group, event or community
- Detailed qualitative data, more insight for further research, allows study of unethical or impractical(S)
- Cant generalise, Researcher bias may occur, hard to replicate and time consuming (L)
54
Observational Design
Behavioural Categories- A behaviour checklist to efficiently see what the researcher sees eg a tally
Event Sampling- Measuring the amount of times a behaviour (the event) occurs in the target pop
Time Sampling- Observing behaviour of target pop in a certain time frame eg every 4 mins
55
Constructing questions
Avoid jargon- use of words not everyone understands
Emotive language- Description which puts researcher view and impacts pressure demand characteristics
Leading questions- Saying a question in a way that leads the person to answer in a certain way
Double barrelled question- 2 questions in one, may agree with half and not the other
Double negatives- Use of more than one negative word influencing the person
56
Role of peer review
- To check research to allocate funding for it if it seems worthwhile
- To validate quality and relevance of the research
- To suggest amendments, if it isn't good enough it can be withdrawn
57
Evaluation of peer review
- Anonymity: Peer reviewers should be kept anonymous however some use this to excessively critique rivals to get funding(thats limited) for their own research. Open review is favoured by some journals
- Publication bias: Publishers prefer showing positive results and might not show other results, inaccurate rep
- Burying research-Peer reviewers may withdraw opposing studies to their own beliefs or mainstream ideas
58
Reliability
Consistency of research (should stay similar if tested again)
59
Internal reliability
Different parts of a test should give the same results
60
External reliability
Test should produce consistent results regardless of when it is used
61
Inter observer/rater reliability
Test should remain consistent regardless of who conducts it
62
Assessing reliability
External, test re-test method- repeat test with same participants, look for 0.8 strong positive correlation
Interobserver- correlate results from each researcher look for 0.8 strong pos correlation.
63
Validity
How well a test measures what it claims to
64
Types of validity
Internal-extent to which results are caused by IV rather than other variables
External- Ecological validity: whether it reflects real life
Population validity:whether it reflects the pop
Temporal-whether findings remain true over time
Face validity- if it seems to measure what it appears to measure
Concurrent validity-If results are similar to an existing test
65
Assessing validity
Internal, face-look at the test to see if it measures what it claims to (eg if questions relate to topic)
Internal, concurrent-compare this test to another existing one, if its valid results will be similar
External, meta-analysis- if results are consistent in different locations its externally valid
External, predictive validity- use data to predict future results, if similar its valid
66
Meta analysis S+L
Systematically combining qualitative and quantitative data to develop an overall conclusion (study of studies)
- Can be generalised to larger pop (S)
- Greater findings can be drawn by correlating results from different studies (S)
- Publication bias: only uses secondary data so not all results which are seen as insignificant may be published(L)
- Researcher bias: Researcher may only use the research which supports their own view (L)
67
Improving reliabiliity
Questionnaires: Check and change questions producing inconsistent data. EG: change open questions to closed, less misinterpretation
- Experiments: Lab exp are more reliable as its controlled
- Interviews: Structured interviews
- Observations: train observers and operationalise categories
68
Improving validity
Questionnaires: Check and change questions producing inconsistent data. EG: change open questions to closed, less misinterpretation. Use lie scales
- Experiments: Lab exp are more reliable as its controlled
- Interviews: Structured interviews
- Observations: train observers and operationalise categories
69
Primary and secondary data S+L
Primary data is collected by the researcher solely for the purpose of their research
- Data can be specific for the aim+ has full control (S)
-Expensive+time consuming
Secondary data is existing info someone else collected
-Time efficient+money efficient(S)
-May not fit criteria of study+variation in quality of data (L)
70
Measures of Central Tendency
Mean- add all no. and divide by amount of no.
Mode-most frequent no.
Median- no. in the middle, if 2 add both and divide by 2
71
Mean, median, mode S+L
Mean: Includes all data, more representative (S), affected by extreme values (L)
Median:Unaffected by extreme values (S), Doesn't work with small data (L)
Mode:Unaffected by extreme values(S), only uses one value (L)
72
Measures of Dispersion
Examines spread of results
Standard deviation:Average amount that all scores deviate from the mean
Range:Minus smallest no. from largest
73
Standard deviation and range S+L
Standard deviation: Uses all of the data (S), time consuming and complex (L)
Range: Quick+easy to cal (S) but doesn't use all data and is affected by extreme values (L)
74
Normal distribution
When the mean median and mode occupy a middle line in the curve- average
75
Skewed distributions
Positive- focused on left and long tail on the right
Negative- focused on the the right and long tail on the left
Mode is always at the peak
Median is always between mode and mean
Mean leans towards the LEFT in NEGATIVE SKEW
Mean leans towards the RIGHT in POSITIVE
this is because it is effected by extreme values+gets pulled to minority result
76
Statistical testing
Shows whether difference or correlation occurs due to chance and how significant it is
An example of stat testing is sign test
77
Systematic sampling
Chosen based on fixed intervals, every nth number
78
Systematic sampling pros+cons
- very easy and can be done manually(S)
- avoids researcher bias(S)
- May only represent certain persona by chance(L)
- Cant represent whole pop
79
Participant variables
Variables in the ppt impacting the DV
80
Types of graphs
- Bar charts*
- Scattergrams*
- Line graphs
- Histograms
81
Bar charts*
- Bar charts should be used when using discrete data (when data is divided into categories + separate. eg: with water + with coffee)
- Categories are on X-axis + frequency/amount of each category on Y-axis
82
Scattergram*
- Scattergrams show associations not differences > used for showing correlations
- One variable on X-axis + the other on Y-axis
83
Histogram
- Bars touch eachother (unlike bar charts) because data is continuous (each number in included without gaps even if its 0)
- X-axis represents equally spaced intervals and Y-axis represents frequency within each interval (eg: marks people get in a test (if 0, keep interval but w/out a bar
84
Line graph
- Represent continuous data + use points connected by lines to show how something changes in value
- IV on X-axis + DV on Y-axis
85
Process of peer review
- Other experienced psychologists in a similar field check + scrutinise the research before it is published
- Look for validity, significance and originality of research
- Assesses how appropriate methods + designs are
- Can suggest adjustments/revisions then re-submit
