Selection Charts

- Name the measure of association and value (Cramer's V = .67).
- Name the strength and, if applicable, direction of the association (e.g. that there is a strong correlation/large effect).
- Repeat the variables to which the measure relates (between the variables 'X' and 'Y').
- Name the units of analysis (e.g. young people).
- Draw a conclusion (the researcher proved to be correct in assuming that...).
- Discuss the percentages from at least two cells in the contingency table if there is an association (10% of the young people who never use WhatsApp watch DWDD, while no fewer than 60% of the young people who often use WhatsApp watch DWDD).

- Name the measure of association and value ((lambda = .35 or: tau = .30, either is sufficient).
- Name the strength and, if applicable, direction of the association (e.g. there is a reasonably strong correlation/medium effect).
- Repeat the variables to which the measure relates (between the variables 'X' and 'Y').
- Name the units of analysis (e.g. women's magazines).
- Draw a conclusion (the researcher proved to be incorrect in assuming that...).
- Discuss the percentages from at least two cells in the contingency table if there is an association (e.g. perfume and cosmetic advertisements in women's magazines make relatively little use of naked men).

Note: if Kendall's tau-b is requested for multiple bivariate correlations, run this via Analyse -> Correlate -> Bivariate -> Kendall's tau-b.

- Name the measure of association and the value (gamma = - .87 or: Kendall's tau-b = - .81, either is sufficient (note the preconditions that apply to the measurements you are using!).
- Name the strength and, if necessary, direction of the association (e.g. there is a very strong negative association).
- Repeat the variables to which the measurement relates (between the variables 'X' and 'Y').
- Name the units of analysis (e.g. Dutch people).
- Draw a conclusion (e.g. the more often X, the less often Y, and vice versa).
- Discuss the percentages from at least two cells in the contingency table if there is an association.

- Name the measure of association and value (rho = .05).
- Name the strength and, if applicable, direction of the association (e.g. there is no association).
- Repeat the variables to which the measure relates (between the variables 'X' and 'Y').
- Name the units of analysis (e.g. Dutch people).
- Draw a conclusion (e.g. there is no association between X and Y).

- Name the measure of association and value (Somers' d = .43).
- Name the strength and, if applicable, direction of the association (e.g. there is a reasonably strongly positive association).
- Repeat the variables to which the measure relates (between the variables 'X' and 'Y').
- Name the units of analysis (e.g. communication science students).
- Draw a conclusion (the more often X, the more often Y).
- Discuss the percentages from at least two cells in the contingency table if there is an association.

- Name the measure of association and value (r = .14).
- Name the strength and, if applicable, direction of the association (e.g. there is a weak negative association).
- Repeat the variables to which the measure relates (between the variables 'X' and 'Y').
- Name the units of analysis (e.g. well-educated women).
- Draw a conclusion (the more often X, the less often Y, and vice versa).

- From all the categories, name the means, the standard deviations and the dependent variable. If comparing more than four categories, it is sufficient to discuss the notable results, such as the highest and lowest means (e.g. on average, people with poor educational qualifications score the highest on the mean number of minutes watching television per week (M = 6.89, SD = 0.78) and the well-educated the lowest (M = 3.56, SD = 1.23).
- Name the measure of association and the value (Eta =.54).
- Name the effect size (e.g. there is a large effect).
- Repeat the variables to which the measure relates (between the variables 'X' and 'Y').
- Name the units of analysis (e.g. Dutch people).

- From all the categories, name the means, the standard deviations and the dependent variable. If comparing more than four categories, it is sufficient to discuss the notable results, such as the highest and lowest means (e.g. on average, people with poor educational qualifications score the highest on the mean number of minutes watching television per week (M = 6.89, SD = 0.78) and the well-educated the lowest (M = 3.56, SD = 1.23).
- Name the measure of association and the value (Eta2 = .29).
- Describe the proportion of declared variance and repeat the variables in it (e.g. 29% of the variance in minutes watching television can be explained by the variance in level of education).
- Name the units of analysis (e.g. Dutch people).

- Name the intercept and the value (a = 8.34).
- Say what this intercept means, repeating the variables (e.g. the model predicts that a person who watches television for zero hours per week reads books for 8.34 hours per week).
- Name the unstandardised regression coefficient and the value (b = -1.73).
- Explain what this b means and repeat the variables (e.g. the model predicts that if the number of hours watching television increases by one per week, the number of hours per week reading books decreases by 1.73).
- State the proportion of declared variance and the value (R2 = .23).
- Describe the proportion of declared variance and repeat the variables (e.g. the variance in the number of hours per week watching television accounts for 23% of the variance in the number of hours per week spent reading books).

- Name the intercept and the value (a = 1.82).
- Say what this intercept means, repeating the variables (e.g. a person who watches television for zero hours per week is zero years old, the model predicts that the person reads books for 5.28 hours per week).
- For each independent variable, name the unstandardised regression coefficient and the values (bwatching television = -0.67 bgender = 1.29; bage = 4.76).
- For each b, explain what this means and repeat the variables (e.g. the model predicts that if the number of hours watching television increases by one per week, the number of hours per week reading books decreases by 0.67).
- If a dichotomous variable (with the values 0 and 1) is used in a multiple regression analysis, the value of the unstandardised regression coefficient is always expressed in the average difference compared to the zero category (e.g. on average, women read books 1.29 hours per week more than men).
- When discussing a regression coefficient , always keep the other independent variables constant (e.g. the model predicts that when the number of hours watching television increases by one hour a week, the number of hours per week reading books decreases by 0.67, if gender and age are constant).
- Say which of the independent variables has the strongest effect on the dependent variable (e.g. gender has the strongest effect on the number of hours per week reading books, Beta = .54).
- Name the proportion of declared variance and the value (R2 = .33).
- Describe the proportion of declared variance and repeat the variables (e.g. the variance in the number of hours per week watching television, gender and age together account for 33% of the variance in the number of hours per week spent reading books).