Skip to content

Advanced Features

Transform (Creating New Variables)

Transform allows you to create new questions (variables) by recoding existing data.

Where to find it

Transform tab → text input field.

Transform Syntax

Each new variable is described by a block of lines separated by an empty line:

New question name
ALT1 {filter_condition} Alternative text 1
ALT2 {filter_condition} Alternative text 2
...
NA {filter_condition}

Where:

  • First line — text of the new question
  • Following lines — recoding rules:
    • ALT — number of the new alternative (integer) or NA (to set "No Answer")
    • {condition} — filter formula in curly braces (same syntax as filters)
    • Text after } — alternative name

💡 If curly braces are empty {}, the condition applies to all questionnaires that have not yet received a value.

Example: recoding from one variable

Age groups
1 {[3]>=18 and [3]<=24} 18-24
2 {[3]>=25 and [3]<=34} 25-34
3 {[3]>=35 and [3]<=44} 35-44
4 {[3]>=45 and [3]<=54} 45-54
5 {[3]>=55} 55+
NA {[3]<18}

Creates a new nominal variable (qn) with 5 alternatives based on metric variable #3 (age).

Example: program sample (Paste Sample)

New Var
1 {[2]=1 or [2]=2} One
2 {[2]=3} Two
3 {[2]=4 and [2]=5} Three
4 {[2]=6} Four
4 {[2]=1} Five
NA {[2]=6}
7 {[2]=9} Seven
8 {} Eight

Important rules

  • If a questionnaire matches multiple conditions, it will receive multiple alternatives → the new variable type will be qj (multi-variant)
  • If each questionnaire matches only one condition — the type will be qn (nominal)
  • The NA {condition} line forcibly sets "No Answer" for questionnaires matching the condition
  • Empty curly braces {} means "everything else" (questionnaires without a value)
  • Multiple transforms are separated by an empty line

Save and load

  • Right-click menu → Save transform — saves text to file
  • Right-click menu → Load transform — loads text from file
  • Right-click menu → Paste sample — inserts a transform example

Recoding Alternatives (Recode)

For quick regrouping of alternatives of an existing variable without writing a transform:

The RecodeIntoNewVar function creates a new variable by combining alternatives of the source variable.

Alternative structure is defined by lines:

1+2
3
4+5+6
7-10
  • 1+2 — alternatives 1 and 2 are combined into one new alternative
  • 3 — alternative 3 remains as is
  • 4+5+6 — alternatives 4, 5 and 6 are combined
  • 7-10 — range of alternatives from 7 to 10

Significance Testing

Significance testing determines whether the difference between values in table columns is statistically significant.

Significance settings

On the Table Settings tab in the Significance section:

Test type

Option Description
✅ Percentage significance Tests difference between percentages in columns
✅ Mean significance Tests difference between means

Comparison

Option Description
With Total Each column is compared with the Total column
With previous column Each column is compared with the previous one
Between columns Columns are compared pairwise (letter notation A, B, C...)

Significance level (Z₀)

Level Z-value
1% Strictest level
5% Standard level
10% Most lenient level

Significance output

Option Description
In Count column Markers added to count cells
In Col% column Markers added to percentage cells
In separate column Separate column for each significance marker

Significance markers

Different symbols can be chosen for marking significant differences:

Style Higher Lower
Triangles ▲ ▼
Arrows ⮝ ⮟
Arrows (Unicode) ↑ ↓
Asterisks ** *

Additional options

  • Use unweighted base — unweighted base (BaseUW) is used for significance calculation

TopBox and BottomBox

TopBoxes allow you to combine top or bottom scale alternatives into one group.

What is TopBox / BottomBox?

  • TopBox — sum of top alternatives (e.g., ratings 9+10 on a 10-point scale)
  • BottomBox — sum of bottom alternatives (e.g., ratings 1+2)

How to add TopBox/BottomBox

  1. On the Table Settings tab in the Boxes section
  2. Right-click
  3. Choose a preset or create your own:
Preset Alternatives
Bottom 1+2 Alternatives 1, 2
Bottom 1+2+3 Alternatives 1, 2, 3
Top 4+5 Alternatives 4, 5
Top 9+10 Alternatives 9, 10
Top 8+9+10 Alternatives 8, 9, 10

Boxes can be added separately to rows (Left) and columns (Top).

Box properties

  • Caption — box name in the table
  • Alternatives — set of PassportValue alternatives included in the box
  • Hide — hide individual alternatives included in the box
  • Boxes participate in mean statistics

Export to SPSS (.sav)

The program allows exporting an FRM array to SPSS .sav format for further analysis.

How to export

  1. Menu FRM → Export to SPSS
  2. Choose encoding:
    • UTF-8 — for modern SPSS versions
    • CP1251 — for older SPSS versions

What is created during export

File Description
array_name.sav SPSS data file
array_name_spss11.mdg MDG file for SPSS 11
array_name_spss12.mdg MDG file for SPSS 12+

Export specifics

  • Metric variables (qm) → SPSS numeric variables (Measure = Scale)
  • Nominal variables (qn) → SPSS numeric variables (Measure = Nominal) with value labels
  • Multi-variant variables (qj) → set of dichotomous variables (v5_1, v5_2, ...) + MDG set
  • Missing value for all variables = -99
  • Variable names: v1, v2, v3... (for qj: v5_1, v5_2, ...)

⚠️ Requirement: SPSS export requires the IBM SPSS I/O Modules library (spssio64.dll / libspssdio.so.1), which must be in the lib/win64/ or lib/lin64/ folder.

Table Merging

Tables can be merged:

Mode Description
No Merging Each table separately (default)
by Column Tables are merged horizontally (by columns)
by Row Tables are merged vertically (by rows)