Home / Resources / Psychological Measures / MCSDS
Marlowe-Crowne Social Desirability Scale:
Scoring and Interpretation
The Marlowe-Crowne Social Desirability Scale (MCSDS) is a 33-item true or false self-report measure of the tendency to give socially approved answers. Marlowe Crowne Social Desirability Scale scoring counts keyed responses and not True answers, and several short forms are not equivalent to the original. This post does not reproduce the item key, so the emphasis falls on how the score is built and used, including when adjusting for it helps and when it removes real variance.
What the MCSDS Measures
Crowne and Marlowe (1960) devised the scale as a measure of social desirability that is independent of psychopathology, as the title of their article states. The items describe behaviors that are culturally approved but uncommon, and shortcomings that are common but denied, so a high score reflects a pattern of presenting oneself favorably.
Researchers use it as a methodological measure in self-report surveys. It is not a detector of dishonesty, and whether a high score reflects a response style or a real personality characteristic is debated, which is why a decision to adjust for it should follow from a stated rationale and not from habit.
Structure of the MCSDS
- Items: thirty-three statements answered True or False. Eighteen are keyed True and fifteen are keyed False (Crowne & Marlowe, 1960).
- Keying: a response that matches the keyed response earns one point, and the opposite response earns zero. Because both keyed-True and keyed-False items exist, a count of True answers is not the score.
- Total: the sum of the 33 keyed indicators, from 0 to 33. The standard form gives one score and no subscales.
- Direction: higher scores mean more socially desirable responding.
- Versions: several short forms exist, and they are not equivalent to the 33-item form. Each has its own key and score range, and this post does not cover them.
- Translations: name the language version and cite validation evidence for it. If none exists, say so and describe how the version was adapted.
The scale was published in Crowne and Marlowe (1960). This post gives no item key, so take the key of the version you administered, check it against the form, and cite it.
How to Score the MCSDS
- Identify the exact version and language administered, and confirm that it is the 33-item form before using a 0 to 33 range.
- Obtain the key of that version, and code each response as 1 when it matches the keyed response and 0 otherwise. Never apply a short-form key to the 33-item form.
- Sum the 33 keyed indicators to obtain mcsds_total, which ranges from 0 to 33.
- Apply a missing-item rule set by the protocol. No developer-endorsed prorating rule was verified, so the code below requires all 33 responses. Report how incomplete responses were handled.
- Keep the raw True and False responses, and create the keyed indicators and the total as new variables.
* Illustrative syntax: mk1 to mk33 hold 1 when the response matches the keyed response and 0 otherwise, created from the key of the administered form.
COMPUTE mcsds_total = SUM.33(mk1 TO mk33).
EXECUTE.
FREQUENCIES VARIABLES=mcsds_total /STATISTICS=MEAN STDDEV MIN MAX.
# data: one row per respondent; items mcsds1 to mcsds33 coded True = 1, False = 0.
# key: the keyed response of each item (1 if True is keyed, 0 if False is keyed), taken from the administered form.
key <- rep(NA_integer_, 33)
stopifnot(!anyNA(key), all(key %in% 0:1))
x <- data[paste0("mcsds", 1:33)]
stopifnot(all(vapply(x, function(z) all(is.na(z) | z %in% 0:1), logical(1))))
keyed <- as.data.frame(Map(function(z, k) as.integer(z == k), x, key))
data$mcsds_total <- ifelse(complete.cases(x), rowSums(keyed), NA_real_)
Interpreting MCSDS Scores
| Score | Interpretation |
|---|---|
| 0 to 33 | Count of keyed responses; higher means more socially desirable responding |
| Study-defined threshold | Use one only if it was validated for your version and population, and prespecify it |
| Diagnostic or clinical cutoff | None exists |
Source: Crowne and Marlowe (1960). No universal threshold was verified.
Higher totals reflect a stronger tendency to endorse socially approved statements and to deny common shortcomings. A score does not prove deception, and it does not diagnose anything. What a given total means depends on the version, the setting and the population, so compare scores only within one form.
Do not replace the key with a count of True answers, and do not apply a short-form key or a published cutoff to the 33-item form. Interpretation must match the version administered.
Psychometric Properties
| Property | Published finding | Source |
|---|---|---|
| Development | A scale of social desirability constructed to be independent of psychopathology | Crowne & Marlowe (1960) |
| Keying | Thirty-three items, 18 keyed True and 15 keyed False | Crowne & Marlowe (1960) |
| Reliability | Estimate internal consistency in your own sample, using a coefficient suited to binary items such as KR-20 | Study-specific |
| Short forms | Each short form needs its own evidence and cannot borrow the properties of the 33-item form | Form-specific |
These are source-specific findings and cannot be assumed to hold as reliability or validity estimates in a new sample, language or form.
Using the MCSDS in a Study
Decide whether the score is a covariate, a moderator, a sensitivity-analysis variable or an outcome, and give the reason. Entering it as a covariate by habit can remove variance that belongs to the construct of interest, particularly when the trait being studied is itself related to favorable self-presentation. A more informative approach is often to report the main result with and without the score and to describe how the estimate changes.
Check the distribution of the total, which is bounded between 0 and 33, and estimate reliability for the sample. Consider how the mode of administration, anonymity and the setting may affect responding, and compare subgroups only after considering whether the measure behaves equivalently across them. Plan sample size from the primary hypothesis and the precision needed, and describe the missing-data method and the covariate decisions in the methods.
Reporting the MCSDS
Socially desirable responding was measured with the [33-item] Marlowe-Crowne Social Desirability Scale (MCSDS; Crowne & Marlowe, 1960) in the [language] version. Each response matching the keyed answer received one point, and points were summed to a total (range = 0 to 33), with higher scores indicating greater socially desirable responding. Missing responses were handled by [method]. Internal consistency in the present sample was [KR-20 = XX], and the score was used as [covariate, moderator or sensitivity variable] in [model].
References
Crowne, D. P., & Marlowe, D. (1960). A new scale of social desirability independent of psychopathology. Journal of Consulting Psychology, 24(4), 349-354. https://doi.org/10.1037/h0047358
Frequently Asked Questions
Is the MCSDS score a count of True answers?
No. Both keyed-True and keyed-False items exist, so each response is scored against the key and the keyed responses are summed.
Does a high score mean a participant is lying?
No. It reflects a tendency toward socially approved answers, and it does not prove deception.
Should I always control for the MCSDS?
No. Adjustment can remove real variance, so use it for a stated reason and consider reporting results with and without it.
Can I use a short form with the 33-item key?
No. Short forms are not equivalent to the original and need their own keys and evidence.
NEXT STEP
Need MCSDS Scores Set Up
or Checked?
OnlineSPSS is a statistical consultancy for researchers. Share your design and dataset, and a statistically trained Research Coordinator reviews it. The scope and one fixed project fee are confirmed before work begins.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014