5
Experimental Research Methods (Subject) / L7 (Lesson)
After this lecture and studying the materials, students are able to:
• Explain the model of a two-way ANOVA
• Interpret the SPSS output of simple effects
• Determine whether a design is balanced or unbalanced
• Explain the advantages of a two-way ANOVA over a one-way ANOVA
• Determine whether a factor is a fixed or random factor
This lesson was created by tklmn.
This lesson is not released for learning.
- simple effects conducting a one-way ANOVA at each level of theother factor A one-way ANOVA with drug as factor for depressed patients A one-way ANOVA with drug as factor for schizophrenic patients
- Balanced designs Equal njk or cell sizes are proportional to each othe All other cases: Unbalanced designs
- Benefits two-way ANOVA over one-way ANOVA (i) Efficiency: 1 instead of 2 experiments(ii) Control: you control for the effect of other variables(iii) Interaction: you can see whether there is an interaction effect(iv) Power: you increase the power of the test of one variable if at least one of the null hypotheses of theadded effects (B, AB) is incorrect (see previous slide)
- Fixed factor Only interested in the selected levels of this factor Levels are not selected from a larger number of levelsthat are also interesting Fixed in our examples: We are only interested in theeffects of these three drugs on behavioralimprovement of these two types of patients
- Random factor Also interested in effects of levels that were notselected Selected levels are a selection from a larger numberof levels that are also interesting Example of random-effects models: Effect of aschoolteacher on performance – in design only a fewteachers or schools, but you also want to generalizeyour results to other teachers and schools that werenot selected
Advertisement