In my field (STEM higher ed research), using the same data collection to answer multiple, RELATED research questions is common practice. Collecting data can be prohibitively expensive and time consuming so you need to make it count, but as Jaspreet Kaur noted, it's also important to avoid diverging too far from a main topic or it will almost certainly reduce your completed response rate (see Tourangeau et al (2000) The Psychology of Survey Response for more info). +1 Samantha Curle -- if you have to leave something out, there's always next time if you still want to pursue that line of inquiry.

Typically, I've found it helpful to determine what modules or survey items I can include based on: (1) its centrality to my primary research agenda, (2) its novelty or importance related to foundational theory/research, and (3) whether or not I can keep the average response time under a certain threshold (depending on what, if any, incentive will be offered). My suggestion: pilot two versions of the survey, one with and one without the additional module or construct, and test how much of an impact it has on response time. Ask your pilot participants what they think of that construct in particular. Maybe even have a few of the non-additional group look over the construct after they are done and ask them if they think it would have been too burdensome or a non sequitur. When in doubt, get feedback.

If you weren't planning on piloting the survey with members of the population you will be studying, then... that's a terrible idea. Pilot it.

Finally, when publishing results, it's an ethical research practice to make sure readers understand when different studies are based on the same data. Easiest solution: give identifying information like when the data was collected and from whom in the methods, and use statements like "Published in [1], other findings from this dataset indicate that..." (side note: also a good way to draw the attention of an interested audience to your related work).

Hope this helps! Good luck!

Hello Umair,

Some may disagree with me (and I wouldn't mind if they did so), but I think it might be a good idea to collect the two sets of data at the one time.

In my view, part of the ethics of survey research involves not burdening respondents any more than is "necessary".

Umair - I think there is no problem in conducting a survey for primary data collection in a marketing research and adding a couple of more constructs for another study and eventually split 1 survey into 2 when it comes to building conceptual framework or testing the reliability and validity of the scale. It depends upon the use of your research design. For example, you can use one cell design (Single situation - Single time period); Two Cell Design or Longitudinal Design or Expost – Facto Design (Single situation; Two period or Two situation; Single period); Three or more Cell Design or Panel Design (Single situation; Three or more Time period); Four Cell Design (Covers two situations at Two time); Matched Stage Design (Data collected from one situation at one time and another situation at another time).

Hi Umair,

While this may be appealing, I would warn against it. From personal experience - your thoughts get muddled and you eventually collect way too much data that you will not be able to process for a long time.

I would suggest focusing on the current study - doing that well, then moving on to a second wave of data collection. Reason for this is - through the first wave of data collection, you'll learn a lot and that may well feed into your original plan of collecting all the data at one time.

Hope this helps :)

Good luck with data collection!

In case all the constructs relate to the same sample,then it might be OK.For example if you are studying the brand image,trust,perception,brand value for a particular brand like Nokia then it is ok to ask questions about all these to teh customers of Nokia.On the other hand if you are judging the pricing constructs of Nokia,again it would be ok to ask the questions to the same set of samples.

I'm coming back to this discussion because I'm currently editing a PhD thesis by publication in which the same data have been used as the basis for four journal articles. It has worked well for three of the articles, but I think the fourth article is too repetitive. So, to a large extent, collecting the same kind of data from the same people has been quite successful.

Furthermore, a colleague and I have engaged in the same kind of activity, but with a different original intention. We collected data from over 800 people and, when writing up the results we found that it was not possible to constrain them to a single article. In fact, we ended up writing three separate articles, all very much distinct from each other - and all currently under review with a single journal.

I appreciate Samantha Curle's perspective about ending up being muddle headed. That could certainly occur, and, indeed, should be avoided. However, I have had different experiences. So, there are two sides to the coin.

In my field (STEM higher ed research), using the same data collection to answer multiple, RELATED research questions is common practice. Collecting data can be prohibitively expensive and time consuming so you need to make it count, but as Jaspreet Kaur noted, it's also important to avoid diverging too far from a main topic or it will almost certainly reduce your completed response rate (see Tourangeau et al (2000) The Psychology of Survey Response for more info). +1 Samantha Curle -- if you have to leave something out, there's always next time if you still want to pursue that line of inquiry.

Typically, I've found it helpful to determine what modules or survey items I can include based on: (1) its centrality to my primary research agenda, (2) its novelty or importance related to foundational theory/research, and (3) whether or not I can keep the average response time under a certain threshold (depending on what, if any, incentive will be offered). My suggestion: pilot two versions of the survey, one with and one without the additional module or construct, and test how much of an impact it has on response time. Ask your pilot participants what they think of that construct in particular. Maybe even have a few of the non-additional group look over the construct after they are done and ask them if they think it would have been too burdensome or a non sequitur. When in doubt, get feedback.

If you weren't planning on piloting the survey with members of the population you will be studying, then... that's a terrible idea. Pilot it.

Finally, when publishing results, it's an ethical research practice to make sure readers understand when different studies are based on the same data. Easiest solution: give identifying information like when the data was collected and from whom in the methods, and use statements like "Published in [1], other findings from this dataset indicate that..." (side note: also a good way to draw the attention of an interested audience to your related work).

Hope this helps! Good luck!