Research Methods Flashcards
What three things does research need to consider?
- Data
- Design
- Difference
What are the four levels of measurement?
- Nominal: Discrete categories
- Ordinal: Related categories
- Interval: Subdivided categories
- Ratio: Proportional categories
What are two design types?
- Independent Groups: Unrelated data
- Repeated Measures: Related data
What statistical test should be used for Independent Groups (unrelated data) with nominal data?
Chi-Squared (X^2)
What statistical test should be used for Independent Groups (unrelated data) with Ordinal data?
Mann-Whitney U
What statistical test should be used for Independent Groups (unrelated data) with Interval or Ratio data?
Unrelated T Test
What statistical test should be used for Repeated Measures (related data) with Nominal data?
Sign Test
What statistical test should be used for Repeated Measures (related data) with Ordinal data?
Wilcoxon’s T
What statistical test should be used for Repeated Measures (related data) with Interval or Ratio data?
Related T Test 11
What statistical test should be used for research with Nominal data and a correlation?
Chi-squared (X^2)
What statistical test should be used for research with Ordinal data and a correlation?
Spearman’s Rho (P=)
What statistical test should be used for research with Interval or Ratio data and a correlation?
Pearson’s R
Two new types of validity
- Face Validity
- Concurrent Validity
Face Validity
whether a test, scale or measure appears to measure what it is supposed to measure ‘on the face of it’. This is done by ‘eyeballing’ the measuring instrument or passing it to an expert to check.
Concurrent Validity
Demonstrated when the results obtained are very close to, or match, those obtained on another recognised test measuring the same concept/construct. This is tested by doing a correlation between the participant scores. If it is strong and positive, it is valid.
How can validity be improved?
- Use a control group
- Standardise procedures and institutions
- Use single blind or double blind procedures
Two types of reliability
- Test-retest reliability
- Interobserver reliability
Test-Retest Reliability
used to assess reliability of questionnaires or psychological tests. Administering same test or questionnaire to same people/person on different occasions. Results obtained should be the same or similar. Two score sets are correlated on statistical test.
Interobserver Reliability
Observations to be conducted by teams of at least two. Observers discuss and argue beforehand their interpretation of behaviour categories. Then watch the same event and record independently. They then correlate the data.
What is a Type 1 Error?
Null hypothesis is jected and alternative is accepted, when the null is ‘true’. E.g. saying a result is significant when it is due to chance.
Often called ‘error of optimists’. Often occurs when significance level is too lenient. Likelihood of making a type 1 error is always the same as significance level.
What is a Type 2 Error?
Null hypothesis is accepted but should have been rejected because the alternative is ‘true’. Often called ‘error of pessimists’. Often occurs when significance level is too stringent.
What are the sections of a Psychological Report?
- Abstract
- Introduction
- Method
- Results
- Discussion
- References
- Appendices
Abstract section of a Report
Summary of the study including aims, hypothesis, method, results, conclusions and implications. Allows the reader to determine if the rest of the report is worth reading.
Introduction section of a report
- Review of previous relevant research
- Should lead logically to research
- Should be like a funnel, start general then more specific
- Ends with researcher starting aim and hypothesis
Method section of a report
Detailed description od what was done. Should be enough detail to be able to replicate. Divided into design, participants, apparatus, procedures and ethics.
Results section of a report
Descriptive statistics: tables and graphs showing frequencies,measures of central tendency and measures of dispersion
Inferential statistics: calculated values,critical values, significance levels. Null hypotheis or alternative hypothesis accepted or rejected
- If qualitative, categories and themes are described with examples.
Discussion section of a report
- Summary: briefly reported in a verbal rather than statistical form
- Relationship to previous research: compared to research in introduction
- Strengths and weaknesses with methodolog and improvements
- Implications for theories and possible real world applications
- Contribution the research has made
- Suggestions for future research
Reference section of a report
Full details of any journal articles, books or websites
Appendices section of a report
Any raw data and material are presented.
Two types of Reference
- Journal Reference
- Book Reference
Journal Reference
Surname, Initial, Initial 2. Surname 2. (year). Title. Journal (Italic). Volume. Edition, page numbers
Book Reference
Surname, Initial 1. (year). Title (Italics). Publisher. Page reference.