ubjective measurements of exercise like an exercise diary is not the most reliable way to measure effectiveness of physical activity (PA) since all self-reports are connected with overestimation. Measuring the effects of PA one should consider duration, intensity level and frequency of the exercise, since there is a dose-response relationship concerning the physiological effects of PA. My opinion is that the questions in the diary should cover these perspectives of PA to be able to analyze the effectiveness of the intervention. A suggestion would be to use an objective measurment as a complement, for example the six-minute walk test. Regards!

Thanks for your responses so far. Well, I am well aware of the issues with subjective measurements but in my particular case there is no other way than using diaries to record the exercise routine. Clearly, I am talking about exercise, which is a different construct than PA (Caspersen, 1985 and other people wrote about it). The objective is to see how well my intervention works, and the dependent variable is the frequency and duration of a particular programme. Well, I however disagree that all subjective measured cause overestimation. There are studies that do not confirm this. Just think of people who forget their activities or those who think PA is equal to exercise (they would give you lower PA rates than you would expect). Objective measures (physical fitness) are in place, but my primary outcome is a different one. I really want to know how to best get accurate diary data. I have been digging in the literature and it seems that people just use diaries.

Why dont you try for a Self prepared/ ready made questionnaire which will record Ex done by patient in previous week/ month..So, that you can correlate it with Ex. Diary too..

Hey,

Maybe this paper can help you.

KInd regard Øystein

Article Do Elite Endurance Athletes Report Their Training Accurately?

Hi Øystein,

thanks for that. Nice paper. Athletes are very different from normal people (I am looking at older adults who are not really active) and hence, your study does not really give me what I want. However, it is a good starting point for further search.

Andre

Hi again,

One of my referebces did something similar in recreational athletes, Borresen, 2006. Take a look at that one:-)

Thanks a lot Øystein and Natalie,

think this gives me something useful.

@ Øystein: could you either upload the article or send it to me? My Uni does not subscribe to the journal.

Thanks.

Her you are:-)

We tend to see 'objective' measures as a gold standard at the moment but we also need to be aware of the pitfalls in all measurement techniques. Bear in mind that if you attempt to 'validate' a self-report measure with an 'objective' measure such as accelerometry or pedometry that such validation will never be perfect as all these measures do not assess physical activity in exactly the same way, they each measure a different aspect of physical activity. So accelerometers would not be able to assess the type of physical activity or anything contextual, as a diary would, and would miss out certain forms of activity depending on the placement of the device on the body. So for example an accelerometer placed on the hip would miss any upper body motion, such as lifting weights or carrying shopping bags. Many accelerometers would have to be taken off to swim. A pedometer merely counts steps but says nothing about the frequency, intensity or duration of physical activity. Like self-report measures, these devices rely on the participant engaging with them appropriately. Also, be aware that physical activity guidelines (e.g. 5 x 30 mins per week) were developed on the basis of self-reported physical activity; guidelines based on accelerometry, when they emerge, may well require less of people. These are a few things I discovered when I undertook a rapid scoping review of literature on physical activity measurement in the spring of this year - please let me know if you want references for these points.

Hi Emma,

thanks for your reply. Yes, I am fully aware of the issues you raised. I had several discussions with a number of people and I also did read a lot. It is very clear that objective and subjective measure are working only when the research question does require the respective measure. In my study I do not need any objective instrument, but I just thought there might be something out there that would have looked at how reliable diaries are. The consensus is now that I will go with the diary (only with the diary) and hope that people filled it up. That is simply as good as it gets. Would be great if you can post the links to the studies here and I will check everything out.

Andre,

You have, of course, hit on the major problem we all face in research--the honesty of subject reports no matter what measurement is used. When we use self-reports in our research we don't pay "incentives", but rather "tracking/reporting fees".  We emphasize and re-emphasize throughout the study that no matter what the subject reports as their activity level or product usage,  the tracking fee is paid just for their information, even if it is "0" or near-"0" for the day.  Of course, they have to be reminded of this throughout the study and always reinforce that they wrote it down and do not compliment or express any praise for high levels or activity and always praise for reporting.  

At the end of the study after subjects have received any payments or test findings for their participation, we ask them to complete an anonymous questionnaire critiquing our performance in the study--a sort of "how did we do?" questionnaire. However, among the critique questions are some questions asking about how candidly they reported their data.  They are reminded not to put their name on the questionnaire and to seal it in an envelope and hand it to the researcher.  Thus, you get more candid responses from this anonymous approach.  However, when the subject hands the envelope to the technician, the technician emphasizes how valuable honest data are and asked if they would volunteer to put their name on the envelope if we give them a non-disclosure statement saying that their data would be completely confidential under HIPAA guidelines.  But you have to make it clear that signing is voluntary  and not pressure them to sign their name and we stress this as strongly as we can. We have found over 90% of subjects volunteer and even when using tracking instead of incentive payments.  You can then link the relevant items in the critique questionnaire to the subject's data.  Not a perfect system, but it often gives on dose-related data that allows for classification into compliant, partially compliant and non-compliant sub-groups.

gil kaats

Hello Gil,

this is indeed a valuable idea. I will note it down and will think about it before I go for my last (third) data collection end of this year. I go get your point and will discuss it further. As you said this is indeed a tricky thing, but we want to address it as good as possible.

Thanks a lot!

Andre