⫸STICKY⫷ What makes you complete an online survey for research?

Tronica

Executive Director
Staff member
Joined
Mar 12, 2002
Messages
4,337
I've been modding this forum for a bit now and I've noticed that some online surveys are much more popular than others. For example, the drugs and personality one is working out particularly well, with BL'ers praising it as per below:

I like how it lets you explain things, unlike other surveys where you really have to ponder your answers so that they don't misinterpret.

As a researcher who conducted online surveys myself, I'm wondering if people can tell me why they like this survey or why they chose to complete it. --- or alternatively, why they started a survey and then left it incomplete... or why they decided never to attempt a survey in the first place.

Eg. is it because of the topic, whether it was interesting to you? because of the tone of writing or the questions? how it was presented? how long it was taking to complete it? the sponsors? or something completely different to that?

Universally, researchers who want to recruit people to their surveys think about how to best attract them and I think it's important to be able to learn from both the good and the bad attempts :)

Fire away folks!
 
I fill them out if they intrest me, if I believe in the research and iv got nothing to do.

For example, with the codeine study its of intrest to me as I use massive amounts of codeine in a week. However with the government pondering banning the substance I fear *any* infomation that comes up will be used against us.

I sometimes leave half way through a study if the questions are reptitive and its starting to take too long. 10min is ok but anylonger and I might get distracted by a dog walking down the street or some internet pornography :p.

Im guessing the studies that do the best are the ones on xtc as its probably one of the most popular drugs while the ones that do the worst are the ones with a strict criteria such as location sensitive.

The top 10 studies in the fourm are

1-Drug Preference and Personality
2-[Canada] North American Opiate Medication Initiative
3-AUSTRALIA: Annual ecstasy user survey 2009
4-Clinical Pharmacology of 1.4mg/kg MDA in Humans (San Francisco, CA)
5-[USA] [Chicago] Healthy men and women aged 18-35 who have used ‘dance’ drugs
6-The Discriminative Effects of Tramadol in Humans (Baltimore, MD)
7-Efficacy Study of Vigabatrin to Treat Meth Dependence(CA, FL, KY, LA, MO, NE, OK, TX)
8- [AUS] Survey on OTC codeine use
9-Optimization of IV Ketamine for Treatment Resistant Depression (NYC)
10-Efficacy of Inhaled Cannabis in Diabetic Painful Peripheral Neuropathy (San Diego)


Im only eligable for 3 of thoes studies. 1 I think iv done, no3 iv done before but dont do anymore because I fear the research will be used against me and same with no8

Topic no1 is broad enough that id doesn't see like it can be used against me, the rest of the topics are specific to America and often to a US state
 
interesting
if any bias is perceived
goal of study/survey
if its not going to take over 10 minutes.
 
I mostly look at them if they apply to me (generally prescription painkiller use), and also the goal of the researcher. For example, ones that are geared toward treatment/understanding of addiction I am much more likely to participate in. I think also giving a small place to explain helps, because there are lots of shades of gray in these black and white surveys.
 
Thanks for your thoughts so far.

I also think there should always be a comments field in an online survey. Usually there is one at the end but I always feel cheated when I get to the end and there is no place for comments as I invariably have some! In those cases I've often gone to the effort of the emailing my comments to the researcher, whilst also saying 'you should have had a comments field!' :)
 
I'll usually do them if they apply to me, in terms of location and subject. For example of the 10 that static_mind listed above, I've only done the personality one and the OTC codeine one. It would be no use doing the ecstacy one as I don't do E.

Sometimes I quit if it's taking too long but usually up to 20 mins is ok. I must say I prefer ones with easier choices, sometimes the ones where you have to grade your responses and you have like 7 options, they can be a bit annoying. I know the researcher is trying to get more accurate answers but it can make it frustrating (I'm talking about the ones where you have a statement and you have to say whether you agree strongly/agree moderately/disagree strongly etc).

Btw Tronica, it seems like everytime I look at your post count it's at exactly 500. What's the deal with that, or am I just going crazy? :)
 
Btw Tronica, it seems like everytime I look at your post count it's at exactly 500. What's the deal with that, or am I just going crazy? :)

Wow. 7 years=500 posts. A lot of lurking :) So we'll make that 501 %)
 
Sometimes I quit if it's taking too long but usually up to 20 mins is ok. I must say I prefer ones with easier choices, sometimes the ones where you have to grade your responses and you have like 7 options, they can be a bit annoying. I know the researcher is trying to get more accurate answers but it can make it frustrating (I'm talking about the ones where you have a statement and you have to say whether you agree strongly/agree moderately/disagree strongly etc).

That's interesting. As a survey designer, I had to weigh up the choices here and I did include a few 'attitudinal' statements that I wanted respondents to rate; I used 7-point scales! A lot of folks use 5-point scales. However there is another option, which I wonder whether people would go for - you can, apparently these days, design your survey with a line from totally disagree to totally agree. On that line is a cursor which the respondent can drag to any point on the line, to indicate their level of agreement. That way you can measure someone's agreement in a very fine grained way (eg. by the millimetre or pixel) but the respondent doesn't see a screen cluttered with options.

When I did my survey, 2 years ago now, the research suggested that this method was not well received by respondents so I went with the old style. But I like the idea of it. What do you think?
 
^How would you collate those responses? Do you start at 0 for totally disagree and +1 for every increment towards totally agree which would be 100? Then you have the problem of not everyone responding the same way - I mean that some will naturally be more conservative with their answers, while some will go for the extremes every time.

Although it's less fine grained, having 3 options (agree, disagree, neither), allows you to get very specific responses because although not everyone will know where they fit on a scale of 1-100, generally people can place themselves in one of those three camps.

A case for good understanding and application of methodology, no? :)


EDIT: By the way, why was it that line method was not well received?
 
I complete a research survey if its relevent to my life(like, its about smoking weed or eating dominoes pizza) or if i want to help out x company because its done good work in the past.

I quit if the questions are boring/repetitive and its taking too long(15mins?). I like surveys with CLEAR questions, none of this 'dont you not know when you wont' bs where i cant figure out if YES means the yes im thinking and so on.
 
Did anyone else read the title of this as "what makes you complete: an online survey for research?"...
and I was hoping to share my deepest feelings of what makes me whole inside
 
^How would you collate those responses?

Well this is an issue. Sometimes you want to just know who agreed or disagreed or were unsure and will collapse them into 3 (as you mentioned, this is the least complicated response set). However by having more categories, you do get to see variation which is interesting - for example, I asked these types of statements in my study and when people were a bit iffy about a statement, heaps more chose 'somewhat agree' or 'somewhat disagree' whereas other statements elicited stronger opinions and people tended towards 'strongly' agreeing or disagreeing. So it gives me information not only about whether they agreed or disagreed, but the strength of that feeling. I usually just present them as graphs.

There is debate about where researchers can treat the 7 point scale as a continuous variable from 1 to 7 in statisticaly analysis. Some people do, but the tests people want to use have assumptions which aren't met by data like this. This is one of the reasons people would like to use the 1-100 line option.

By the way, why was it that line method was not well received?

One reasons was that people doing the online survey found it difficult or unusual and were more likely to quit! We need to make surveys likeable, easy, enjoyable, otherwise people will quit. Another reason is that the specificity may just be too much. Eg. one study measured the same variable with the two different scales (7-point versus the line) and found the extra possible answers available via the line method didn't really help the analysis in the end. The debate will continue though and I imagine the line method may become more popular as internet survey respondents might warm to the method. and it is likely to be appropriate for some measuring tasks and not others. Software to create them will become more popular too.
 
1. What's in it for me if I do the survey?
2. What's in it for the rest of our society?

If those two add up to be generally good, and the survey doesn't take longer than 10-15min, I'm in. I would also consider my personal info and the way it's used.
 
This thread has been useful so far. However I've just been thinking that it would be good to have an FAQ in Drug Studies about the good and bad of online surveys - aimed at researchers who hope to attract our audience to their surveys. This thread, and the comments in other DS threads, can feed into that FAQ. I can then send a link of the FAQ to researchers I know.

What do you reckon?

Another question is this: how important is the topic of the survey to you? Many so far have said that interest in the topic (and a sense that the topic is one that will have a positive influence on drug use/harm either for the individual or generally) is critical.

Often I have colleagues who get funded to find out more about treatment needs in the community and look to recruit non-treatment-going drug users to see what their needs are. I find these surveys don't do so well with attracting people's interest - even though the surveys are interested in getting views from people who don't want or feel they need (or actually need) treatment.

What are your thoughts on that?
 
The institution for which it is being composed.

And what type of institutions would put you off or would attract you to do a survey?

(I think I have an idea but don't want to assume)
 
The most important thing to me is if I'm eligible for it.

I've only stopped taking a survery after beginning it if I feel that it's directed in such a way that it will inevitably favour a certain conclusion. I like the outcome of the survey to be open.
 
And what type of institutions would put you off or would attract you to do a survey?

(I think I have an idea but don't want to assume)

If it is for a gonvernment institution, I probably will choose to not, for obvious reasons of distrust.

If it is for an institution of higher learning or psychological/medical foundation, I will.

Also, if it is simply composed by a student to aid in an essay or project, I probably will not, since they are not in such learned position to be composing surveys.
 
At least they are learning.. gotta start somewhere!

I enjoy filling in surveys as they make me analyse my personality in an objective manner.. often come away thinking 'how can i change this'! It's also interesting to see what other people's research designs are like, and what other researchers are interested in at present.
 
Top