Research Methods⚗️ Flashcards

(60 cards)

1
Q

History of Psychological science

A

Interest in human thought throughout history

Philosophy- no evidence, answering theoretical questions is difficult
19C Germany-empirical methods, paradigm shift from religion to science

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

4 key features of science

A

Determinism- systematic causes
Parsimony- simplest cause
Systematic empiricism-structured and organised
Testability- can be falsified using research techniques

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Benefits of science

A

Systematic collection of data otherwise difficult to obtain by observation

Evidence supports answers and rules out others. Theories formed, validity

Improve the world through changes in policy

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Research process

A

Generate research question-concepts, constructs of abstract ideas
Theory-general principle to explain phenomenon
Hypothesis- prediction tested in study
Study- identify population, recruit sample, Identify variables, design and measures
Collect data-ethical principles
Analyse data and interpret results - qualitative/quantitative
Answer research question-draw conclusions about population

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Theory

A

Set of principles that explain a phenomenon, articulated factors and processes in a structured way e.g. flow chart

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Theory’s role in the research process

A

Develops research question (if no empirical evidence, limited evidence or competing explanations)
Answer research question, provide rationale (draw on theory to explain other phenomena)
Results inform theory (disconfirm existing principles or add new ones)
Informs chance in practice and policy

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Why rationales are needed

A

Strongest argument for hypothesis

  • research is costly in time and resources
  • research must contribute to literature
  • cannot simply rely on previous research studies
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

3 types of rationales

A

Novel contribution to conceptual understanding- first evidence for causes and processes, tests competing explanations, replicate results exactly or conceptually

Novel contribution to methodology-improves materials and procedure

Novel contribution to practice-treatment and intervention

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Two types of research question

A

Difference between groups (compare scores of groups)

Relationship between variables (see if individual’s scores on variables are related)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Null hypothesis

A

No effect

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Experimental/alternative hypothesis

A

An effect

Non directional- difference between groups, relationship between variables

Directional-lower/higher scores between groups, positive/negative relationship between variables

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Variation

Systematic and unsystematic

A

Participant responses vary, different scores in same group

Systematic- scores systematically differ (trend in variation) due to effect being investigated

Unsystematic-score variation due to error or random effect other than what is being investigated. Research cannot control it, always present

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Aim and logic of null hypothesis testing

A

Unsystematic variation only= null hypothesis (no effect)

Logic of science, disconfirmation and falsifiability
If no evidence for null then conflicted support for H1 (systematic variation)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Probability

A

Extent to which we were correct to discredit the null

0-1

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Populations

A

Set of individuals with characteristics under investigation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Samples

A

Selected individuals from the population to take part in study

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

Why samples are needed

A

Entire population cannot be sampled
Lack access to resources and time
Use results to make inferences about the population

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

Representative and not representative samples

A

Representative- unbiased, matches phenomenon in society

Not representative-biased, results differ from phenomenon in society

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

Sampling error

A

Difference between results found in sample and in population, how representative it is

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

What is sampling error influenced by

A

Sampling strategy

Size of sample

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

Sampling strategies

A

Random sampling-individuals randomly selected from list of entire population at same,e size

Stratified sampling-individuals randomly selected from relevant subgroups of the population

Opportunity sampling- individuals self select to take part in study to sample size

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

Categorical and continuous variable

A

categorical- Often IV, no numerical value, individual assigned to single category or group e.g. gender

Continuous-Often DV Scores can be ordered on a continuum e.g IQ

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

Experiments to determine differences between groups

A

Between participants- each participant to one condition (independent measures)compare scores between groups

Within participant-each participant in all conditions (repeated measures)scores between individual

True experiment- randomly assign individuals to IV, equal chance of being selected in any condition

Quasi experiment- non random assignment to condition, uses existing groups e.g. gender

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Experiments to determine relationship between variables

A

Correlational- relationship between two continuous variables, linear (positive/negative)

Categorical- relationship between two categorical variables

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Approaches to time
Cross sectional- conducted at a single point in time Longitudinal-conducted over a period of time
26
How to address boredom/practice effects
Counterbalancing
27
Constructs
How abstract ideas can be investigated, may not be visible or have standard exemplars
28
Variable
Operationalised construct, concrete measurement e.g. a scale
29
Measure
How the variable can be assessed in the study
30
5 types of measures
Self report- verbal statements, response scales tell researcher directly Implicit measures-participant demonstrates, reaction times etc, automatic Biological measures-not consciously controlled, cortisol levels etc, compare to standard Task performance-score on recall task etc, shows researcher Behaviour-keep track of who does what, observed frequency
31
Finding a measure
Already published Develop new measure if not studied yet Scores should directly link to variable, low demand characteristics
32
3 research settings
Field studies-real world settings, relevant to phenomenon, event already occurring. Researcher has limited control Lab studies- random allocation, artificial environment may lack ecological validity. Less confounding variables Online/mail studies-participant chooses setting within timeframe. Measure outcome variables. Experimental manipulation with text and images
33
Ethics historical context
Participants first seen as subjects, little consideration Nazi research alerted world to participant experience Nuremberg code developed to protect all humans, inspired modern ethics
34
Risks of research
Physical, psychological harm (anxiety ,shame) Waste time for little purpose Misinterpret or misuse research
35
How to gain approval for research
Follow the code to publish research, declare all principles are followed Approval granted by ethics committee before conducting study-British Psychological Society
36
BPS 4 general principles and the ethical considerations
Respect autonomy, privacy and dignity of individuals and communities (informed consent, freedom from coercion, confidentiality, anonymity, debriefing) ``` Maximise benefit, minimise harm (minimise physical/psychological harm) Scientific integrity (contribute knowledge and insights) ``` Social responsibility to individuals and communities (avoid misuse and misinterpretation of results)
37
Types of deception
Passive- withhold full truth, key pieces of relevant info. Specific variables and hypothesis not mentioned Active- intentionally misinform participants about the study e.g. confederates
38
Why is deception used
To study actual responses to real events Telling participants everything could influence their responses or cause demand characteristics May not be able to recreate a real world event while telling the full truth Passive is more acceptable but active maximises experimental realism and researcher control
39
Ethical issues with deception
Violates informed consent May experience harm, distress when told about deception May mistrust researchers and authority
40
Why we need to quantify variation
To differentiate between systematic and unsystematic variation Is not valid/reliable to estimate sizes Use descriptive statistics
41
Descriptive statistics
Summarise individual’s responses Characteristics of a sample Measures of central tendency and dispersion
42
Measures of central tendency and what measures they are used for
Score most representative or typical Mean- continuous Median-continuous Mode- categorical
43
Mean
Uses all available data and reflects actual value of all scores Can be influenced by outliers
44
Median
Less affected by outliers Does not use the value of all the scores
45
Measures of dispersion
Spread or distribution of scores. Allows variation in scores to be quantified Range Variation Standard deviation
46
Range
No information in variation of other scores in the set
47
Variance
Distribution of scores around the mean Calculate how much each score deviates from the mean Square each value and add all up Divide by frequency MINUS ONE
48
Standard deviation
Take the square root of the variance Original measurement units
49
How to calculate the lower and upper range using SD
Lower range 1SD - mean Upper range 1SD + mean (Can be 2SD, 3SD...)
50
The 68-95-99.7 rule
Percentage of all included scores are consistent across all normal distributions 68% scores lie within 1SD of the mean 95% scores lie within 2SD of the mean 99.7% scores lie within 3SD of the mean
51
How to predict scores in a population
Mean +/- 1[or 2 or 3] SD
52
How to predict the population mean Issues with this
Sample recruited to estimate population parameters (mean/SD) May have sampling error or unsystematic variation CALCULATE A CONFIDENCE INTERVAL
53
Confidence intervals
Range of scores likely to indicate the true population mean Typically focus on 95% confidence Uses the 69-95-99.7 rule to determine how many SDs around the sample mean will yield 95% confidence
54
Confidence interval formula
95% uses 1.96SD not 2SD to have perfect distribution Mean +/- 1.96 x SD divided by the square root of n N(sample size)
55
How does increasing the sample size give a more precise estimate
Gives a smaller interval size | Reduces sampling error and less discrepancy between the sample and the population
56
Trade off between uncertainty and precision, why 95% is used Certainty definition
High certainty and high precision (smaller interval size) Certainty-level of confidence that true population mean lies within the range estimated from the sample
57
The p value
Uncertainty, how likely to make a type 1 error Each null has a p value to indicate the chance of this Can never be 100% certain about our inferences
58
Decision error
When sample statistics do not match the real population, unsystematic variation Incorrect decision to accept or reject the null
59
Type 1 and 2 errors
Type 1 error- find effect when none is there, null rejected when should have been accepted Type 2 error-find no effect when there is one, null accepted when should have been rejected
60
Interval size
Range of scores in which the population mean lies Smaller interval size means a more precise estimate