How to do the test:

  1. State your Null Hypothesis glossary item: Null Hypothesis in the form:
    "The dependent variable is not related to the independent variable in a linear fashion"
  2. Choose a Critical Significance Level (α: alpha) glossary item: Critical Significance Level
    This is typically α = 0.05
  3. Calculate the test statistic
    Use a spreadsheet to calculate your results, or
    Access a helpsheet for SPSS
    and check here for data entry into SPSS
  4. Reject or accept your Null Hypothesisglossary item: Null Hypothesis
    Either compare the calculated statistic with Critical Value, glossary item: Critical Value
    If F ≥ Fcritical → reject H0 → significant result
    If F < Fcritical → accept H0 → non-significant result
    Look at the calculated probability (P value) glossary item: probability
    If P ≤ α → reject H0 → significant result
    If P > α → accept H0 → non-significant result
  5. Report your results
    Plot your data using the appropriate graphs
    (Regression test: Fdf regression, df error = …, P = …)
  6. If your regression test is significant, you can use your model for prediction glossary item: Prediction, using a linear model glossary item: Linear model of the form:
    y = bx + c
    (where y is the dependent variable glossary item: Dependent variable,
    x is the independent variable glossary item: Independent variable,
    b is the slope of the line glossary item: Slope and
    c is the constant or intercept glossary item: Constant)
    Use the coefficient of determination, R2 glossary item: Coefficient of determination, to indicate how much of the variability in y is explained by the regression
    View an example report