How to Revise Inferential testing — AQA A-Level Psychology
Inferential testing constitutes the mathematical framework for determining the probability that observed data patterns are attributable to chance rather than experimental manipulation. It requires the application of the null hypothesis significance testing (NHST) paradigm, necessitating the selection of statistical tests based on measurement levels (nominal, ordinal, interval) and design types (related, unrelated). This process is fundamental to establishing the scientific validity of psychological claims and ensuring the robustness of conclusions regarding human behavior within the global research community.
Examiner Tips for Inferential testing
- Use a decision tree or mnemonic to help select the correct statistical test based on the three criteria: difference vs correlation, experimental design, and level of measurement
- Always check if the hypothesis is one-tailed or two-tailed when looking up critical values in statistical tables
- Remember that for most tests, the calculated value must be equal to or greater than the critical value to be significant (the exception being the sign test, Wilcoxon, and Mann-Whitney where the calculated value must be equal to or less than the critical value)
- Practice calculating the sign test as it is a foundational requirement
Common Mistakes in Inferential testing
- Confusing Type I and Type II errors
- Incorrectly identifying the level of measurement (nominal, ordinal, interval)
- Failing to correctly identify the experimental design (repeated measures, independent groups, matched pairs) when selecting a test
- Misinterpreting the relationship between the calculated value and the critical value (e.g., whether it needs to be greater than or less than the critical value for significance)
Key Marking Points
- Knowledge of the sign test (when to use and calculation)
- Understanding of probability and significance levels (p=0.05)
- Use of statistical tables and critical values to determine significance
- Distinction between Type I and Type II errors
- Factors influencing the choice of statistical test (level of measurement and experimental design)
- Knowledge of specific tests: Spearman’s rho, Pearson’s r, Wilcoxon, Mann-Whitney, related t-test, unrelated t-test, and Chi-Squared test