Note
Studied by 5 people
Untitled Flashcards Set
WEEK 1: APPROACHES TO RESEARCH DESIGN
Q: What are the two main types of research methods?
A:
Quantitative Methods: Test theories using numbers.
Qualitative Methods: Analyse behaviour and language (e.g., observations, conversations).
Q: What are the steps in the research process?
A:
Initial observation.
Theory generation.
Hypothesis formation.
Data collection.
Data analysis.
Q: What are the types of measurement?
A:
Categorical: Divides entities into distinct categories (e.g., cat vs dog).
Continuous: Produces scores (e.g., amount an animal eats).
Q: What is measurement error?
A: The discrepancy between actual values and represented values.
Q: What are the two main types of research design?
A:
Non-Experimental: Observes natural behaviours without interference.
Experimental: Manipulates variables to examine effects on outcomes.
Q: What are the two types of study designs?
A:
Between Groups: Different entities in different experimental conditions.
Within Groups: Same entities participate in all conditions.
DESCRIPTIVE STATISTICS
Q: What are the measures of central tendency?
A:
Mean: Average of values.
Median: Middle value in an ordered set.
Mode: Most frequent value.
Q: What are the types of data based on modes?
A:
Bimodal: Two modes.
Multimodal: Multiple modes.
Q: How can data distribution be visualised?
A: Using histograms, which display frequency distributions.
NORMAL DISTRIBUTION
Q: What are the types of skewness?
A:
Positive Skew: Scores cluster at the lower end; tail points to high values.
Negative Skew: Scores cluster at the higher end; tail points to low values.
Q: What are the types of kurtosis?
A:
Leptokurtic: Positive kurtosis (peaked).
Platykurtic: Negative kurtosis (flat).
Q: How can normality be checked?
A:
Histograms.
MMM alignment (Mean, Median, Mode).
Skew ≈ 0, Kurtosis ≈ 0.
Trimmed mean analysis.
SCREENING DATA
Q: What is homoscedasticity?
A: Variance is consistent across all levels of an independent variable.
Q: What is heteroscedasticity?
A: Variance is inconsistent across levels of an independent variable.
Q: What are Z-scores?
A: Standardises data so the mean = 0 and standard deviation = 1.
BASICS OF STATISTICAL ANALYSIS – CHOOSING THE RIGHT TEST
Q: What should be identified in a research question?
A: Determine independent variables (IV) and dependent variables (DV); decide if they are categorical or continuous.
Q: What are null and alternative hypotheses?
A:
Null Hypothesis (H₀): No effect or difference exists.
Alternative Hypothesis (H₁): A significant effect or difference exists.
Q: What is statistical significance?
A: The likelihood of a result occurring by chance, indicated by a p-value.
CORRELATION
Q: What does correlation measure?
A: The relationship between two variables.
Q: What are the assumptions for correlation?
A:
Linearity.
Normality.
Homoscedasticity.
No extreme outliers.
Q: What are common non-parametric tests for correlation?
A:
Spearman’s Rho: Used for ranked data.
Kendall’s Tau: Used for small samples.
Q: What is the difference between partial and semi-partial correlation?
A:
Partial Correlation: Controls for a third variable affecting both variables.
Semi-Partial Correlation: Controls for a third variable affecting only one variable.
EFFECT SIZE
Q: What is effect size?
A: A measure of the strength of the relationship between variables.
Q: What are benchmarks for Pearson’s r?
A:
Small: ±0.1
Medium: ±0.3
Large: ±0.5
Q: What is r²?
A: The proportion of variance shared by two variables.
T-TESTS
Q: What are the types of t-tests?
A:
Independent Samples T-Test: Compares two groups based on independent data.
Paired Samples T-Test: Compares two related groups (e.g., pre- and post-test).
Q: What are assumptions of t-tests?
A:
Normal distribution of scores.
Homogeneity of variances (tested using Levene’s test).
ANOVA (Analysis of Variance)
Q: What is the purpose of ANOVA?
A: To compare means of more than two groups.
Q: What are the types of ANOVA?
A:
One-Way ANOVA: Single independent variable.
Two-Way ANOVA: Two independent variables.
Factorial ANOVA: Examines interactions between multiple independent variables.
Q: Why not use multiple t-tests?
A: Increases Type I error rates; use Bonferroni corrections instead.
ANCOVA (Analysis of Covariance)
Q: What is the purpose of ANCOVA?
A: Extends ANOVA by controlling for extraneous variables.
RESEARCH ETHICS
Q: What are the key ethical principles?
A:
Respect: Protect participants' dignity and rights.
Beneficence: Minimise harm and maximise benefits.
Justice: Fair distribution of research burdens and benefits.
Q: What are important historical guidelines?
A:
Nuremberg Code (1949): Informed consent and voluntary participation.
Declaration of Helsinki (1964): Emphasised risk-benefit analysis.
Belmont Report (1979): Introduced principles of respect, beneficence, and justice.
Q: What are examples of unethical research?
A:
Milgram Obedience Studies (1961): Deception and emotional harm.
Stanford Prison Experiment (1971): Psychological harm to participants.
Henrietta Lacks (1951): Cells taken without consent.
FACTORIAL DESIGNS
Q: What are factorial designs?
A: Research designs that include two or more independent variables.
Q: What are main and interaction effects?
A:
Main Effect: Independent impact of one variable on the DV.
Interaction Effect: Combined effects of IVs on the DV.
Q: What are examples of factorial designs?
A:
2x2 Design: Two IVs, each with two levels.
Example: Alcohol consumption and gender on attractiveness.
REGRESSION
Q: What is regression?
A: A method for predicting the value of one variable (DV) based on one or more predictor variables (IVs).
Q: What is the regression equation?
A:
Y=b0+b1X1+b2X2+ᴈY=b0+b1X1+b2X2+ᴈ
b0b0: Intercept.
b1b1: Slope.
ᴈᴈ: Error term.
Q: What are assumptions for regression?
A:
Linearity.
Independence of observations.
Homoscedasticity.
Multicollinearity (VIF < 10).
MIXED ANOVA
Q: What is Mixed ANOVA?
A: A combination of within- and between-groups designs.
Q: What is an example of Mixed ANOVA?
A: Comparing diet types (between) and time (within) on weight loss.