Written and maintained by CASRAI Editorial Board
Last updated
Who this is for: researchers and students who need to score, interpret or cite the International Physical Activity Questionnaire (IPAQ) and know how much weight its numbers can carry.
What is the IPAQ?
The IPAQ is a self-report questionnaire that estimates how much physical activity a person did, in walking, moderate-intensity and vigorous-intensity activity, over a recall window. It was developed by an international consensus group in 1997-98 as four long and four short forms, each in telephone or self-administered format and using either a “last 7 days” or “usual week” reference period (Craig et al., 2003). Its output is either a continuous score in MET-minutes per week or a three-level category: low, moderate or high.
What is an iPAQ device?
An iPAQ (with a lowercase i) is an unrelated handheld computer brand. This page covers only the questionnaire, which is written IPAQ.
What is a physical activity questionnaire?
A physical activity questionnaire asks respondents to report the type, frequency and duration of activity over a stated period, so that activity can be estimated without wearing a device. Because the answers are recalled and self-reported, they are subject to the biases covered in social desirability bias, and their accuracy has to be judged against an objective criterion such as an accelerometer. The IPAQ is one widely used example; the CDC youth survey described in the YRBS guide is another instrument that asks about physical activity.
How to score the IPAQ
Formula
The official scoring page gives the continuous score as MET level × minutes of activity × events per week, computed for each activity type and then summed (IPAQ Research Committee, scoring page, read 2026-11-07). The short-form validity literature writes this as:
Total MET-min/week = 8.0 × vigorous + 4.0 × moderate + 3.3 × walking
where each of “vigorous”, “moderate” and “walking” is minutes per day multiplied by days per week (Lee et al., 2011). A MET (metabolic equivalent of task) is a multiple of resting energy expenditure, so 8.0 means about eight times rest. The full data-processing protocol (revised November 2005) also defines how to clean implausible or very short responses; apply it from the official protocol before analysis.
Worked example
Suppose a respondent reports vigorous activity 30 minutes on 3 days, moderate activity 45 minutes on 2 days, and walking 20 minutes on 5 days.
- Vigorous: 30 × 3 = 90 min; 8.0 × 90 = 720 MET-min
- Moderate: 45 × 2 = 90 min; 4.0 × 90 = 360 MET-min
- Walking: 20 × 5 = 100 min; 3.3 × 100 = 330 MET-min
- Total: 720 + 360 + 330 = 1,410 MET-min/week
Category: the respondent did vigorous activity on at least 3 days at 20 or more minutes a day, which meets the moderate criterion. The high criteria need either vigorous activity on at least 3 days totalling 1,500 MET-min/week (here only 720) or any combination on 7 days totalling 3,000 MET-min/week (here 1,410). The category is therefore moderate.
Category decision table
| Category | Criteria (from the official scoring page) |
|---|---|
| High | Vigorous activity on 3 or more days totalling at least 1,500 MET-min/week, OR any combination of walking, moderate and vigorous activity on 7 days totalling at least 3,000 MET-min/week |
| Moderate | Any one of: vigorous activity on 3 or more days at 20 or more minutes a day; moderate activity or walking on 5 or more days at 30 or more minutes a day; any combination on 5 or more days totalling at least 600 MET-min/week |
| Low | Does not meet the criteria for the moderate or high categories |
Calculating it in Excel, R or Python
Excel, with vigorous minutes × days in B2, moderate in C2, walking in D2:
=8*B2 + 4*C2 + 3.3*D2
R:
met_min <- function(vig, mod, walk) 8 * vig + 4 * mod + 3.3 * walk met_min(90, 90, 100) # 1410
Python:
def met_min(vig, mod, walk):
return 8 * vig + 4 * mod + 3.3 * walk
print(met_min(90, 90, 100)) # 1410.0
How to report it (APA)
Example sentence: “Physical activity was assessed with the IPAQ short form and scored as MET-min/week following the IPAQ scoring protocol (Craig et al., 2003).” Reference entry: Craig, C. L., Marshall, A. L., Sjöström, M., Bauman, A. E., Booth, M. L., Ainsworth, B. E., Pratt, M., Ekelund, U., Yngve, A., Sallis, J. F., & Oja, P. (2003). International physical activity questionnaire: 12-country reliability and validity. Medicine & Science in Sports & Exercise, 35(8), 1381-1395. https://doi.org/10.1249/01.MSS.0000078923.96621.1D
Reliability and validity of the 12-country IPAQ study
In the 2000 study across 14 centres in 12 countries, test-retest reliability (Spearman ρ) clustered around 0.8, and the short and long forms performed comparably. Criterion validity against a CSA (now MTI) accelerometer had a median ρ of about 0.30, which the authors described as comparable to other self-report validations. They concluded the IPAQ had acceptable measurement properties for monitoring population levels of physical activity among adults aged 18 to 65 in diverse settings, and recommended the short form with “last 7 days” recall for national monitoring and the long form for research needing more detail (Craig et al., 2003).
A later systematic review of 23 short-form validation studies found correlations with objective measures ranging from 0.09 to 0.39, below the minimum acceptable standards used in that review (0.50 for activity monitors, 0.40 for fitness measures). In the six studies that compared absolute levels, the short form usually overestimated activity, by 36 to 173 percent, and the authors concluded that evidence supporting the short form as an indicator of relative or absolute activity was weak (Lee et al., 2011).
How to interpret IPAQ scores
| Evidence | Value | Reading |
|---|---|---|
| Test-retest reliability (Craig et al., 2003) | ρ about 0.8 | Responses are fairly repeatable across time |
| Criterion validity vs accelerometer (Craig et al., 2003) | median ρ about 0.30 | Weak-to-modest agreement with measured movement |
| Short-form validity range (Lee et al., 2011) | 0.09 to 0.39 | Below the review’s minimum acceptable standard |
| Absolute level (Lee et al., 2011) | overestimated by 36 to 173% in most studies | Do not treat MET-min values as measured energy expenditure |
The gap between high reliability and modest criterion validity is the point to remember: a questionnaire can give consistent answers that still diverge from measured activity. For that distinction see convergent and discriminant validity. Scores bunched at the top or bottom of the scale raise the issues in floor and ceiling effects.
Assumptions and common mistakes
- Using the MET values (8.0, 4.0, 3.3) without stating them; report them.
- Comparing scores from different recall windows (“last 7 days” versus “usual week”) as if they were the same measure.
- Treating MET-min/week as precise or as a measure of fitness; the validation evidence above shows it is a rough self-report estimate.
- Applying the adult 18-65 validation to other groups without separate evidence. The official download page lists an elderly version and translations, and states it cannot check the accuracy of the translations (read 2026-11-07).
- Skipping the official data-processing protocol and inventing your own cleaning rules.
FAQ
How do I interpret the IPAQ scale score?
Use the category (low, moderate, high) from the table above, or the MET-min/week total for continuous analysis. Treat it as a relative, self-reported estimate, and note the overestimation found in validation reviews.
Which form should I use, short or long?
Craig et al. (2003) recommended the short form with 7-day recall for national monitoring and the long form where more detailed information is needed.
Where do I get the questionnaire and scoring protocol?
From the IPAQ website: download page (questionnaires) and scoring page (protocol), both read 2026-11-07.
Sources
- IPAQ Research Committee. IPAQ scoring page and Guidelines for Data Processing and Analysis (revised November 2005). https://sites.google.com/view/ipaq/score (read 2026-11-07)
- Craig, C. L., et al. (2003). Med Sci Sports Exerc, 35(8), 1381-1395. PMID 12900694.
- Lee, P. H., Macfarlane, D. J., Lam, T. H., & Stewart, S. M. (2011). Int J Behav Nutr Phys Act, 8, 115. https://doi.org/10.1186/1479-5868-8-115
