Basic statistical analysis of the European Social Survay 2018 data.
Data were taken from the ESS survey: https://www.europeansocialsurvey.org/
Data analysis was carried out at different levels of measurement: nominal, ordinal, quantitative, and quotient.
(a) Nominal/ ordinal x nominal/ ordinal variable
Analysis steps: Cross-tabulation with column percentages; test of independence, i.e. chi-squared test, bootstrapped Chi2, Fisher's exact test; a measure of the strength of association, mosaic or bar/column plot by category of the other variable.
Description of variables:
- ordinal variable called "psppipla" - degree of conviction about the impact on the political system (Not at all; Very little; Some; A lot; A great deal)
- nominal variable called "vote01" - did the person vote? (YES, NO)
(b) Nominal/ ordinal x quantitative variable
Descriptive statistics (means, variances, IQR, etc.) by group by category of the second variable; test of independence, i.e. t-student, ANOVA, regression or their non-parametric equivalents; a measure of the strength of association: R2 from regression; box plot or density plot or histogram of a quantitative variable by grouping by category of nominal/order variable
Description of variables:
- independent variable votes: nominal variable called "vote01" - whether the person participates in the election
- dependent variable police: quantitative variable called "trstplc" - how much does the person trust the police (NO 0-10 complete trust)
(c) Quantitative x quantitative variable
Descriptive statistics (mean, variance, IQR, etc.) of both variables; Pearson correlation or R2 from regression; scatter plot with a line of best fit
Description of variables:
- independent variable comp: quantitative variable called "netustm" - how much a person uses the computer (in minutes)
- dependent variable confidence: quantitative variable called "pplhlp" - how much do you think that others are likely to try to be helpful, or are they only concerned about themselves? (people only care about themselves 0-10 people are helpful)
Project was done for the Statistics course 2022/2023 at the Jagiellonian University.