ERSS 6824
SUMMARIES
REMEMBER:
In “data view”:
o Each column = variable
o Each row = respondent
Distinguish between data levels, because different descriptive statistics are used:
CATEGORICAL CONTINUOUS
Nominal Interval
Age can be categorical or nominal
Ordinal Ratio
How to know: in variable view, it will show How to know: in variable view, it will not
values show any values
Frequencies Descriptives
No reliability analysis Reliability analysis
Biographical data is mostly categorical Scales, ex. Likert scales
QUESTIONS THAT SHOULD BE ASKED WHEN ANALYSING ANY DATASET:
1. How do I determine the frequencies of my biographical data?
o Code the data according to biographical information
o Use descriptive statistics (mean, mode, skewness, standard deviation, etc.)
2. How do I determine the psychometric properties of my questionnaire?
o Conduct reliability analysis (Cronbach’s alpha)
o Conduct factor analysis
3. How do I determine if there is a relationship between certain variables?
o Use correlation analysis (Pearson’s Product Moment Correlation / Chi-Square test
of independence)
4. How do I determine which variables (independent variables) are significant predictors
of an outcome (criterion)?
o Multiple regression
5. How do I determine if there are any significant differences between two groups?
o If the sample is small and skewed, use Mann-Whitney U test
o If the sample size is relatively big, use t-test
6. How do I determine if there are any significant differences between amongst 3 or
more groups?
o If the samples is small and skewed, use Kruskal-Wallis test
o If the sample is relatively big, use ANOVA
, HOW TO OPEN:
File – open – data – look for drive – open/double click
SAVE
Only ONE data file
Only ONE output file
HOW TO KNOW THE MEANINGS OF VALUES UNDER EACH VARIABLE:
In variable view: values –
SUMMARIES
REMEMBER:
In “data view”:
o Each column = variable
o Each row = respondent
Distinguish between data levels, because different descriptive statistics are used:
CATEGORICAL CONTINUOUS
Nominal Interval
Age can be categorical or nominal
Ordinal Ratio
How to know: in variable view, it will show How to know: in variable view, it will not
values show any values
Frequencies Descriptives
No reliability analysis Reliability analysis
Biographical data is mostly categorical Scales, ex. Likert scales
QUESTIONS THAT SHOULD BE ASKED WHEN ANALYSING ANY DATASET:
1. How do I determine the frequencies of my biographical data?
o Code the data according to biographical information
o Use descriptive statistics (mean, mode, skewness, standard deviation, etc.)
2. How do I determine the psychometric properties of my questionnaire?
o Conduct reliability analysis (Cronbach’s alpha)
o Conduct factor analysis
3. How do I determine if there is a relationship between certain variables?
o Use correlation analysis (Pearson’s Product Moment Correlation / Chi-Square test
of independence)
4. How do I determine which variables (independent variables) are significant predictors
of an outcome (criterion)?
o Multiple regression
5. How do I determine if there are any significant differences between two groups?
o If the sample is small and skewed, use Mann-Whitney U test
o If the sample size is relatively big, use t-test
6. How do I determine if there are any significant differences between amongst 3 or
more groups?
o If the samples is small and skewed, use Kruskal-Wallis test
o If the sample is relatively big, use ANOVA
, HOW TO OPEN:
File – open – data – look for drive – open/double click
SAVE
Only ONE data file
Only ONE output file
HOW TO KNOW THE MEANINGS OF VALUES UNDER EACH VARIABLE:
In variable view: values –
