Exam 1

  1. Metainformation

    Tag Value
    fileInferential_Statistics_uu-Oneway-ANOVA-851-en_uu-Oneway-ANOVA-851-en
    nameuu-Oneway-ANOVA-851-en
    sectionInferential Statistics/Parametric Techniques/ANOVA/Oneway ANOVA
    typeschoice
    solutionFALSE, FALSE, FALSE, TRUE
    TypeCase
    Program
    LanguageEnglish
    LevelStatistical Reasoning

    Question

    Four education science students conduct a study for the bachelor’s thesis on the effects of concept mapping on learning performance of psychology students (N = 86). The study distinguishes three conditions:

    1. Pure: Students receiving instruction in concept mapping. 2. Supportive: Students receiving instruction in creating the concept map, with part of the concept map already given. 3. Control: Students who do not receive instruction in creating the concept map.

    Four weeks after the psychology students read the chapter, they take a knowledge test (ten questions where each correct answer earns 1 point). Students in experimental conditions 1 and 2 are expected to score higher on average on the knowledge test than students in the control condition.

    The education science students analyze the data with an ANOVA. They note the following data from this in their report: F(2,83)=17.90F(2, 83) = 17.90, p<.001p < .001, η2=0.13\eta^2 = 0.13.

    Which of the following statements is correct based on the reported test result?


    1. FALSE: The means of experimental conditions 1 and 2 are significantly higher than the mean in the control condition.
    2. FALSE: The effect of condition is significant but not relevant (less than 5% explained variance).
    3. FALSE: The reported data may not be from this study.
    4. TRUE: Post hoc analysis is needed to determine whether in experimental conditions 1 and 2 the means are significantly higher than in the control condition.

    Solution

    This study compares 3 different conditions. Finding a significant difference between these groups (p < .001), we do not yet know which of the 3 groups differ from each other. This requires a post hoc analysis.