SPSS command: SPSS ORTHOPLAN

Orthogonal designs for conjoint studies

An orthogonal design picks a small fraction of all attribute combinations so that every attribute's main effect can still be estimated independently.

Run Orthogonal Design (ORTHOPLAN) in your browser All SPSS procedures

Free · No account · Files stay on your device · Updated 18 October 2026

What it tells you

Crosstabs runs the SPSS ORTHOPLAN procedure in your browser: describe the factors and levels, and it builds the plan cards, adds holdout cards and verifies orthogonality.

How many cards respondents must rate instead of the full factorial.
The attribute levels on each card, ready to turn into profiles or a questionnaire.
Whether every pair of factors is orthogonal and whether every level appears equally often.

When to use it

Use it when

Planning a full-profile conjoint study: price, brand and feature attributes with two to nine levels each.
The full factorial is too large for one respondent to rate (for example 3 × 3 × 3 × 2 × 2 = 108 profiles).
You want holdout cards to test how well the estimated utilities predict new profiles.

Use something else when

You need interactions between attributes: a main-effects plan cannot estimate them.
Choice-based (discrete choice) conjoint with choice sets: this generates single profiles for rating or ranking, not choice tasks.
Some level combinations are impossible and must be prohibited: ORTHOPLAN has no prohibitions, so edit the plan or use another design method.

How to run it in Crosstabs

01
Open Statistics mode
Open any file in the workspace and switch to Statistics mode. Orthogonal design does not read the rows of the open file.
02
Choose Orthogonal design
In the Classify group, pick Orthogonal design (ORTHOPLAN).
03
Describe the factors
In the Factors table give each factor a Name, a Label and its Levels (comma separated), 2–9 levels each; use Add factor and Remove factor. At least two factors are needed.
04
Set cards and seed
Set Minimum number of cards (0 lets the plan decide), Holdout cards (default 4) and Random seed (default 2000000), then click Generate orthogonal design.
05
Download or analyse
Click Download cards (CSV) to get one row per card with STATUS_ and CARD_ columns. In the same session the plan is also offered to Conjoint analysis.

Run Orthogonal Design (ORTHOPLAN) in your browser

Worked example

Computed by the Crosstabs engine from the synthetic data below when this page was built — not typed in.

Data

A meal-kit subscription study with four attributes: price per box (49, 59, 69), menu style (Classic, Vegetarian, Family), delivery (Weekly, Fortnightly) and trial offer (None, First box free). Minimum number of cards 0 (let the plan decide), 4 holdout cards, random seed 2000000 — the panel's defaults apart from the factors. No data file is needed.

Orthogonal Design

Card	Price per box	Menu style	Delivery	Trial offer	Status
1	69	Vegetarian	Weekly	None	Design
2	69	Family	Weekly	First box free	Design
3	49	Family	Weekly	None	Design
4	69	Classic	Fortnightly	None	Design
5	59	Family	Fortnightly	None	Design
6	59	Vegetarian	Weekly	None	Design
7	49	Vegetarian	Fortnightly	First box free	Design
8	59	Classic	Weekly	First box free	Design
9	49	Classic	Weekly	None	Design
10	69	Classic	Weekly	None	Holdout
11	69	Vegetarian	Weekly	First box free	Holdout
12	69	Vegetarian	Fortnightly	First box free	Holdout
13	49	Family	Fortnightly	None	Holdout

Base plan PB3.4 (9 runs); seed 2000000. Orthogonality check passed: every pair of factors has proportional level frequencies. Some factors have unequal level frequencies (levels were collapsed); orthogonality is unaffected.

How to read it

9 design cards replace the 36 combinations of the full factorial, and every pair of factors is orthogonal, so each attribute's main effect can be estimated independently of the others. 4 holdout cards are added: respondents rate them like any other card, but CONJOINT leaves them out of the estimation and uses them to check predictive validity. The two-level factors come from collapsed three-level columns, so their levels are not equally frequent — orthogonality still holds.

Output tables explained

Plan cards: A summary of how many design cards replace the full factorial and how many holdout cards were added, with a warning when no library plan reaches the requested minimum.
Orthogonal Design: One row per card with the level of every factor and its Status (Design or Holdout); footnotes name the base plan, the seed, and the orthogonality and balance checks.

Assumptions and notes

Only main effects are estimable: the plan assumes attributes contribute additively to preference.
Collapsing a three-level column to two levels keeps orthogonality but makes level frequencies unequal; the footnote says when this happens.
The same factors, minimum and seed always give the same cards, so a plan can be regenerated exactly.

How the results are validated

Tests check that every plan in the built-in library — each Plackett–Burman plan, each fixed plan and each column split — is an orthogonal array, that the SPSS decision rules choose the expected base plan for common designs (the 3 × 3 × 3 × 2 × 2 carpet-cleaner design gives 16 cards), that 300 random designs are all orthogonal and meet the requested minimum whenever the library allows, and that a seed reproduces the same cards and CSV layout. Validation evidence

Questions

What makes a design orthogonal?

For every pair of factors, the levels occur together in proportion to their separate frequencies, so knowing one factor's level tells you nothing about another's. That is what lets each main effect be estimated independently.

Why did I get more cards than the minimum?

Orthogonal arrays exist only for certain numbers of runs. The panel picks the smallest library plan that fits your factors and the minimum; replicated base plans are reduced to one copy of each distinct card.

What are holdout cards for?

Respondents rate them like other cards, but CONJOINT does not use them to estimate utilities. The correlation between their observed and predicted preferences (Kendall's tau for holdouts) checks predictive validity.

Can I change the cards after generating them?

Changing a level on a design card breaks orthogonality. Regenerate with another Random seed instead, which reassigns factors to columns, or edit only holdout cards.

How do I collect the ratings?

Turn each card into a profile in your survey tool and record one column per card in card order, one row per respondent. That layout is what the Conjoint analysis panel reads.

Run Orthogonal Design (ORTHOPLAN) on your own data

Open a .sav, CSV or Excel file. Everything runs in your browser; the file never leaves your device.

Open the workspace

SPSS is a trademark of IBM Corp. Crosstabs is not affiliated with or endorsed by IBM.