Reasearch Methods Part 3 Flashcards
Theory Construction
An educated assumption that is constructed through direct observation
Hypothesis testing
Experimentally checking if the hypothesis is valid
Paradigms
Shared assumptions and understandings within a scientific discipline that has set laws
Paradigm shift
Significant change in the underlying theory within scientific discipline
Objectivity
We minimise the influence of personal bias. This maintains validity and limits EV’s
3 ways to increase objectivity
○ Discussing and comparing results
○ Including a 3rd party
○ Blinding the researcher from the aim
Replicability
Extent to which scientific findings and research can be repeated
Which research methods are easy to replicate?
Lab, controlled observations, Questionnaire
Which research methods are difficult to replicate?
Field, natural, naturalistic observation, unstructured interview
Null hypothesis
Assumes there is NO relationship between the two variables
Falsifiability
The theory is not scientific until it can be disproved
Abstract
Summary of all sections
Max. 200 words
Idea whether report is relevant to researcher
Introduction
Literature review, which summarises relevant theories and justifies the research. Initially broad but then becomes increasingly specific. Ends with aims and hypotheses
Method
Allow others to replicate the study. Includes:
○ Ppts
○ Design
○ Materials
○ Procedure
Results
Accepts/rejects null hypothesis. Brief summary of findings in written form
Quantitative studies - Descriptive and inferential statistics
Qualitative studies - thematic analysis
Discussion
Summarises findings in the context of the research. Discusses potentially limitations and how other researchers can reduce bias. Practical applications and implications are highlighted and conclusions are made
References
Organised in alphabetical order
○ Credits other researchers
○ Avoids plagiarism
○ Allows readers to find materials you’ve used
Define case studies
Extremely in depth investigations of:
○ A single individual - unique circumstances
○ A group - demographics
○ An institution
Two types of case studies
Snapshot - looks at behaviour over a short period
Longitudinal - looks at behaviour over a longer period, sometimes years or decades
Example of a case study
Genie
○ She was kept away from society and researchers investigated how she would develop her language and behaviour when exposed to wider society
Example of a case study
Genie
○ She was kept away from society and researchers investigated how she would develop her language and behaviour when exposed to wider society
Strengths of Case studies
Rich, detailed data - could be missed if controlled research methods used
Conduct research that would otherwise be unethical
Weakness of case studies
Low population validity - experience of one individual may not be generalisable
Ethical issue - identity may not remain confidential due to uniqueness of individuals case
Define content analysis
Analysising media
Qualitative -> quanitative
Newspapers, social media, interviews
Define content analysis
Analysising media
Qualitative -> quanitative
Newspapers, social media, interviews
How do you carry out content analysis
1) Create categories
2) Read through data
3) Tally the number of times each category features
Define thematic analysis
Analysing media
Qualitative -> qualitative
Picking out common themes
How do you carry out thematic analysis?
1) Create categories (coding)
2) Read through data
3) Create themes that summarise emergent ideas
Strengths of content analysis
Easily replicated - communication sources can be analysed by other researchers
High ecological validity - sources are real-life situations
Define reliability
How consistent something is
Test-Retest
Test ppts using a questionnaire, then after some time retest them using the same questionnaire
Correlate the results, and the results should have a correlation coefficient of 0.8
Inter-observer
Several observers record their results separately and then correlate the results, and they should get a positive correlation coefficient of 0.8
Interal validity
Did the research measure what it intended to
○ High internal validity - good level of control over EV. IV is likely to have influenced DV
○ Low internal validity - EV’s have confounded the results. Variables other than IV have influenced DV
External validity
Are findings from the research generalisable
○ Temporal validity
○ Population validity
○ Ecological validity
Concurrent validity
1) Get ppts to complete standardised questionnaire
2) Ask ppts to complete the new questionnaire
3) Correlate results, correlation efficient of 0.8
Face Validity
Does the questionnaire do what it says it does
Nominal data
○ Discrete categories
○ Mode
Ordinal data
○ Ordered data
○ Not equal intervals
○ Likert scale
○ Mode or median
Interval data
○ Natural order
○ Equal intervals
○ Mean, median, mode
Difference, unrelated, nominal
Chi-squared test
Difference, unrelated, ordinal
Mann-Whitney U
Difference, unrelated, interval
Unrelated t-test
Difference, related, nominal
Sign test
Difference, related, ordinal
Wilcoxon
Difference, related, interval
Related t-test
Correlation, nominal
Chi-squared test
Correlation, ordinal
Spearman’s Rho
Correlation, interval
Pearson’s R
Inferential statistics
Whether results from sample apply to target population
Type 1 error
Optimism error:
Alternative hypothesis accepted incorrectly
Falsley stating results are significant
May be due to using a more lenient level of significance
Type 2 error
Pessimism error:
Alternative hypothesis rejected incorrectly
Falsely stating results aren’t statistically significant
May be due to using stringent level of significance
Sign test
Add new column
Calculate difference and place +/-
Add +/- total
Less frequent sign used
Degrees of freedom
Chai-squared test
(r-1) x (c-1)
R = row
C = column
