Thank you very much Kathrin. Will give it a read now.

Dear Saima,

These papers maybe help you.

Saint-Maurice, P. F., Welk, G. J., Beyler, N. K., Bartee, R. T., & Heelan, K. a. (2014). Calibration of self-report tools for physical activity research: the Physical Activity Questionnaire (PAQ). BMC Public Health, 14(461), 1–9. http://doi.org/10.1186/1471-2458-14-461

Vanhelst, J., Béghin, L., Turck, D., & Gottrand, F. (2011). New validated thresholds for various intensities of physical activity in adolescents using the Actigraph accelerometer. International Journal of Rehabilitation Research. Internationale Zeitschrift Fur Rehabilitationsforschung. Revue Internationale de Recherches de Readaptation, 34(2), 175–177. http://doi.org/10.1097/MRR.0b013e328340129e

Sabia, S., van Hees, V. T., Shipley, M. J., Trenell, M. I., Hagger-Johnson, G., Elbaz, A., … Singh-Manoux, A. (2014). Association between questionnaire- and accelerometer-assessed physical activity: the role of sociodemographic factors. American Journal of Epidemiology, 179(6), 781–90. http://doi.org/10.1093/aje/kwt330

Nang, E. E. K., Gitau Ngunjiri, S. A., Wu, Y., Salim, A., Tai, E. S., Lee, J., & Van Dam, R. M. (2011). Validity of the International Physical Activity Questionnaire and the Singapore Prospective Study Program physical activity questionnaire in a multiethnic urban Asian population. BMC Medical Research Methodology, 11, 141. http://doi.org/10.1186/1471-2288-11-141

Puyau, M. R., Adolph, A. L., Vohra, F. a, Butte, N. F., Maurice, R., Vohra, A., & Validation, N. F. B. (2002). Validation and calibration of physical activity monitors in children. Obesity Research, 10(3), 150–157. http://doi.org/10.1038/oby.2002.24

van Hees, V., Fang, Z., Langford, J., Assah, F., Mohammad, A., da Silva, I., … Brage, S. (2014). Auto-calibration of accelerometer data for free-living physical activity assessment using local gravity and temperature: an evaluation on four continents. Journal of Applied Physiology, 44(0), 738–744. http://doi.org/10.1152/japplphysiol.00421.2014

many thanks Abdel-Rahman :)

Have you considered that if your final goal is to estimate correlations of physical activity with other variables (e.g., health outcomes), you don't need to calibrate your questionnaire scores? By "calibrating" questionnaire scores with accelerometer data beforehand, then generating correlations with secondary variables, you may actually lose important information.

Dear David,

Thank you for your answer. Can you please help me understand further what important information I will be missing please?

Best 

Saima

Saima, any time you use a measure, scores reflect various things. They are systematically affected by whatever construct the test measures, sometimes systematically affected by other unintentional constructs (such as social desirability, education level, etc.), and randomly affected by various sources of error.

There is an assumption that accelerometers are a gold standard measure of "physical activity", but they only measure certain aspects of physical activity, particularly the gross bodily movement aspect.

Questionnaires also measure physical activity. They capture how much gross bodily movement occurs over the timeframe (as reflected by personal perceptions of movement), and may also capture other types, or elements, of physical activity that accelerometers don't capture very well (such as flexibility training, strength training, cycling, rowing, etc.) 

Regressing PA questionnaire scores on accelerometer data only focuses on equating the PA questionnaire scores with whatever the accelerometer captures - therefore any PA that is captured by the questionnaire that *isn't* typically captured by the accelerometer is lost.

I hope that helps to at least partly explain what I was getting at.

Hello Saima, 

I have some questions for you:

1) Do you need to validate your questionnaire? Or Is it already validated (e.g. GPAQ, IPAQ?

2) What was your method? Did you apply the questionnaire before or after accelerometer measures?  

I was reading the answers that You could have and I think that David answer its quite exact. Besides, I'm also working with accelerometer (Actigraph GT3X+) and I think that you must pay attention to different cut points suggested by Actigraph. I mean, If You use Freedson's (1998) one there are 60 epoch as reference to adults. This reference can affect your interpretation of data. There is an important impact of accelerometer epochs on classification activity patterns.

I hope that my comments will be useful for you. 

Have a nice time.

Hi Gonzalo,

1) I want to validate them in a specific population. Although these have been validated against gold standards in other populations.

2) My participants wore the monitor for 7 days and then at the end of the 7 days they filled these 7 day recall questionnaires.

I have used Freedson (2011) cutpoints.

So, Should I skip the calibration step and just find the association of physical activity measured by these questionnaires with metabolic parameter?

Actually one of these questionnaires was used in a prospective study where they only used a questionnaire and not the accelerometer. So I thought I will calibrate this questionnaire and then use the prospective data. (I hope I make some sense)

Thank you for your time

Saima

Dear David,

Thank you very much for helping me understand better.

Saima

Hello Saima,

1) I want to validate them in a specific population. Although these have been validated against gold standards in other populations. A.1. I think that you can validate your questionnaire doing convergent validity If there are others questionnaires for this population. In this case you apply exitent questionnaire and yours at the same time. If everything is ok,  you would have the almost the same factors.

2) My participants wore the monitor for 7 days and then at the end of the 7 days they filled these 7 day recall questionnaires. A2. I suggest you to apply your questionnaire before accelerometer because you could know the predictive capacity of your questionnaire to measure Pyysical activity. Or you could have a bias may be accelerometer could affect their perception of daily physical activity.

I have used Freedson (2011) cutpoints. A3. The cutpoint is ok.

So, Should I skip the calibration step and just find the association of physical activity measured by these questionnaires with metabolic parameter? A4. Calibration step could be after convergent validity to make your questionnaire more robust.

Actually one of these questionnaires was used in a prospective study where they only used a questionnaire and not the accelerometer. So I thought I will calibrate this questionnaire and then use the prospective data. (I hope I make some sense) A5. As I said in my A1. You could use this questionnaire to make a convergent validity.

Have a good time. Do not hesitate to contact me.

Gonzalo

Calibration of physical activity questionnaire using Accelerometer? - ResearchGate. Available from: https://www.researchgate.net/post/calibration_of_physical_activity_questionnaire_using_Accelerometer [accessed Aug 15, 2016].