One of key features of Analyze are broad analytical capabilities resulting, among others, from availability of various calculations. All calculations that can be found in Analyze are listed below with short explanations. Very important thing about calculations in Analyze is that they are always performed on weighted values (more details about setting a weighting variable in: Selection of weighting factor).
Group
For numerical or categorical questions?
Statistics
Statistics
Numerical and Categorical (converted from Numerical)*
Mean – arithmetical average of weighted values.
Minimum – smallest value from a set of data.
Maximum – largest value from a set of data.
Variance – the average of the squared differences from the Mean weighted by selected variable.
Std deviation – squared root of variance / average difference from a mean value weighted by selected variable.
Median – value that splits an ordered set of data for two parts where 50% of values is smaller and 50% is greater than median.
First quantile – in a set of data, a number for which 25% of data is smaller and 75% is higher than that number.
Third quantile – for a set of data, a number for which 75% of data is smaller and 25% is higher than that number.
Mode – most frequent value.
Counts and totals
Numerical and Categorical
Unweighted count – unweighted number of responses/cases (excluding missing values).
Weighted count – number of responses/cases weighted by selected variable.
Unweighted total sum – sums unweighted values.
Weighted total sum – sums weighted values.
Percentages
Categorical
% of column, sample base – share of given answer in a column sample total.
% of row, sample base – share of given answer in a row sample total.
% of table, sample base – share of given answer in a column and row sample total.
% of column, subset base – share of given answer in a column subset total.
% of row, subset base – share of given answer in a column subset total.
% of table, subset base – share of given answer in a column and row subset total.
Statistics below can be used to calculate % share of answers where the base is an entire branch of nested questions.
% of column total, sample base – share of given answer in a whole column sample base.
% of row total, sample base – share of given answer in a whole row sample base.
% of table total, sample base – share of given answer in a whole column and row sample base.
% of column total, subset base – share of given answer in a whole column subset total.
% of row total, subset base – share of given answer in a whole row subset total.
% of table total, subset base – share of given answer in a whole column and row subset total.
Top-N-Box
Numerical and Categorical (converted from Numerical)*
Share of Top-N or Bottom-N most desired answers in answers overall. Please keep in mind that Top-N-box calculations are available only for questions for which it is reasonable from analytical perspective to calculate this statistic.
% of Top-1-Box
% of Top-2-Box
% of Top-3-Box
% of Top-4-Box
% of Top-5-Box
% of Bottom-1-Box
% of Bottom-2-Box
% of Bottom-3-Box
% of Bottom-4-Box
% of Bottom-5-Box
* This calculations group can be applied only for numerical or numerical converted to categorical questions. Still if question was converted to categorical you can use these statistics only in Subset total and Sample total calculations. More details about question conversion in Conversion: numeric or categorical?
