- Messages
- 26
@ Site Controllers
This is about the Diabetes Survey which your email wanted me to contribute to. If this is the wrong place, I apologise but on looking at the (over-busy) home page I didn't find anywhere else to write this. If it should go elsewhere, please would someone give me an address.
Would I like to win an iPad? - yes. And help with Diabetes Research? Naturally.
Take a survey - only 2 minutes. 2 minutes? - check
Name, age, location, date of diagnosis. All of these I can do.
Control of symptoms - I take statins for high cholesterol, so I assume they have some impact on diabetes as well. Or does this only apply to those whose doctors have specifically diagnosed statins for diabetes? Hm - decide cholesterol control close enough to diabetes to include them.
Do I use any other injectable medicines? Well, yes I do. I use Copaxone - a disease-modifying drug for MS. Does this count on a diabetes-related questionnaire? Guess - probably not. Of course if you're after data about general health, or about problems which arise from injections generally, then it might be relevant. Still - let's say no. Note, however, that a phrase such as "for your diabetes" might have helped. Many of us have problems which do not arise from diabetes.
In my case treatment is supposed to be careful control of diet. Maybe there will be questions later about how carefully I control my diet - and I will even try for honesty.
Then on to blood glucose testing
How often do I check my blood glucose? I don't - my GP said that it wasn't necessary in a very discouraging sort of tone. (Would I like to - well, actually I 'd like to do so for a while so that I can get a feel for how diabetes works in my case, but probably not regularly, not yet.)
How often would I like to test? See temporarily, whilst that is an option as above.
Have I been refused test strips - well, sort of - see "GP says don't do it" above.
Was I refused after having been prescribed them - well, no - see above
Which blood test meter do I use? EUREKA! An answer which I can answer fully and honestly. An option which says, "I do not use one."
Next question, "Is your meter bought privately or prescribed?" Un-Eureka, de-Eureka. Two options only, radio buttons and if you try to click on Next and ignore the question, you're out of luck. You must pick one. Which should I pick, given the previous answer?
Now I could have lied and kept going in the hope of winning an iPad, but I'm not that mercenary, or that optimistic, or that free with my time given that the data will be worse than useless. I'm giving up instead - at least you won't get fictitious data.
So I'm going to give you some genuine feedback, and some advice:
Ideally a survey should filter the questions according the answers given in previous questions - see the online version of the 2011 Census as an example of how to do this trick.
If you want everyone to answer every question, then you should at least include a "not applicable" option. It may be that my GP (and the practice's "lead diabetic nurse") are so far behind/ahead of the times that their regime is one the survey framers haven't come across, but it seems unlikely.
This survey is badly designed.
When thinking about the survey don't just work out what information you want, work out what information your target might be giving. If what you are actually trying to do is get some statistics on the vexed question of prescribing test strips, then say so, and let people for whom this is a non-issue, because they are not trying to test, opt out of your survey earlier. Perhaps question X, "Do you do regular bg testing at home? Yes/No" question X+1, "Have you ever done bg testing at home" Yes/No question x+2 "Do you want to do bg testing at home?" Then if you get 3 "No" answers in a row skip to the next issue.
Test a survey with less likely answers as well as with the predictable ones - or tell people who don't fall into the predictable range not to bother.
Don't assume answers to questions you didn't ask. So far it looks like you know that , "Everyone would like to test bg if cost weren't an issue." So you keep on about cost. Well, it may be an issue for those who would like to test or are testing, but in my case my GP saying, "Don't do it." is a bigger one. May be I will over-ride her, maybe not.
It may be that some of these possibilities, and other issues, are covered later in your survey. I don't know, I've given up part way through because, after several questions which were difficult to answer accurately, I finally got through to one which was impossible to answer accurately in this format. I had little confidence that the rest of the survey was going to be any better, and decided to explain this as my contribution to the project.
I'm not knocking the idea of this survey, but truly you haven't thought this through and/or haven't tested enough. A badly designed survey may give you enough data to use for a PR campaign, but not if you're after real information.
This is about the Diabetes Survey which your email wanted me to contribute to. If this is the wrong place, I apologise but on looking at the (over-busy) home page I didn't find anywhere else to write this. If it should go elsewhere, please would someone give me an address.
Would I like to win an iPad? - yes. And help with Diabetes Research? Naturally.
Take a survey - only 2 minutes. 2 minutes? - check
Name, age, location, date of diagnosis. All of these I can do.
Control of symptoms - I take statins for high cholesterol, so I assume they have some impact on diabetes as well. Or does this only apply to those whose doctors have specifically diagnosed statins for diabetes? Hm - decide cholesterol control close enough to diabetes to include them.
Do I use any other injectable medicines? Well, yes I do. I use Copaxone - a disease-modifying drug for MS. Does this count on a diabetes-related questionnaire? Guess - probably not. Of course if you're after data about general health, or about problems which arise from injections generally, then it might be relevant. Still - let's say no. Note, however, that a phrase such as "for your diabetes" might have helped. Many of us have problems which do not arise from diabetes.
In my case treatment is supposed to be careful control of diet. Maybe there will be questions later about how carefully I control my diet - and I will even try for honesty.
Then on to blood glucose testing
How often do I check my blood glucose? I don't - my GP said that it wasn't necessary in a very discouraging sort of tone. (Would I like to - well, actually I 'd like to do so for a while so that I can get a feel for how diabetes works in my case, but probably not regularly, not yet.)
How often would I like to test? See temporarily, whilst that is an option as above.
Have I been refused test strips - well, sort of - see "GP says don't do it" above.
Was I refused after having been prescribed them - well, no - see above
Which blood test meter do I use? EUREKA! An answer which I can answer fully and honestly. An option which says, "I do not use one."
Next question, "Is your meter bought privately or prescribed?" Un-Eureka, de-Eureka. Two options only, radio buttons and if you try to click on Next and ignore the question, you're out of luck. You must pick one. Which should I pick, given the previous answer?
Now I could have lied and kept going in the hope of winning an iPad, but I'm not that mercenary, or that optimistic, or that free with my time given that the data will be worse than useless. I'm giving up instead - at least you won't get fictitious data.
So I'm going to give you some genuine feedback, and some advice:
Ideally a survey should filter the questions according the answers given in previous questions - see the online version of the 2011 Census as an example of how to do this trick.
If you want everyone to answer every question, then you should at least include a "not applicable" option. It may be that my GP (and the practice's "lead diabetic nurse") are so far behind/ahead of the times that their regime is one the survey framers haven't come across, but it seems unlikely.
This survey is badly designed.
When thinking about the survey don't just work out what information you want, work out what information your target might be giving. If what you are actually trying to do is get some statistics on the vexed question of prescribing test strips, then say so, and let people for whom this is a non-issue, because they are not trying to test, opt out of your survey earlier. Perhaps question X, "Do you do regular bg testing at home? Yes/No" question X+1, "Have you ever done bg testing at home" Yes/No question x+2 "Do you want to do bg testing at home?" Then if you get 3 "No" answers in a row skip to the next issue.
Test a survey with less likely answers as well as with the predictable ones - or tell people who don't fall into the predictable range not to bother.
Don't assume answers to questions you didn't ask. So far it looks like you know that , "Everyone would like to test bg if cost weren't an issue." So you keep on about cost. Well, it may be an issue for those who would like to test or are testing, but in my case my GP saying, "Don't do it." is a bigger one. May be I will over-ride her, maybe not.
It may be that some of these possibilities, and other issues, are covered later in your survey. I don't know, I've given up part way through because, after several questions which were difficult to answer accurately, I finally got through to one which was impossible to answer accurately in this format. I had little confidence that the rest of the survey was going to be any better, and decided to explain this as my contribution to the project.
I'm not knocking the idea of this survey, but truly you haven't thought this through and/or haven't tested enough. A badly designed survey may give you enough data to use for a PR campaign, but not if you're after real information.