Notice: the WebSM website has not been updated since the beginning of 2018.

Two blogs of the month, October 2012.

Dec 04 2012

For this month, we present you the two blogs of month, instead of just one. For the first blog of the October, we have chosen How to minimize memory errors in surveys from the Relevant Insights blog.

The article in blog is talking about thinking of past events or past behavior, of which you may have been asked in any kind of survey. Michaela Mora, the author of the blog claims, that the questions about past are quite common in the surveys and they can be problematic. As she say: "Respondents, often and unintentionally make memory-based errors which lead to over- or understatements of their actions.".

There are a lot of factors, that can affect the memory of the survey respondent:

  • The type of event is being asked about
  • Feelings about and meaning of the event
  • How often an event has happened
  • How long ago the event happened
  • How the question is formulated

Those factors are responsible for our ability to recall the particular event and they are responsible for so called telescoping effect. The author also talks about the difference between the episodic memory and semantic memory.

If we want tominimize those errors, based on the memory recall, we can follow those recommendations for the survey and question design:

  • Keep the reference period for the event of interest closer to the time in which the data is collected.
  • Refer to meaningful events to anchor the time frame.
  • Provide clues that can be associated with the behavior of interest
  • Use “warming-up” questions
  • Refer to previous answers
  • Use records if available

For more detailed information on the recommendations above and the article itself, you can read the whole article by following the link above.

For the second blog of the month, we are presenting the article called Radical data collection practices for the "smart" world. This article was posted on the Kinesis Survey blog.

As the Kinesis Survey team claims: "Strategy Analytics recently reported that global smartphones in use have now surpassed 1 billion, and will reach 2 billion by 2015.", so we at the WebSM.org are thinking the same. This is why we have also chosen this article as the second blog article for this month. They have previously recommending the quick and easy modifications for the mobile surveys, mainly for the market researchers, but we think that their recommendations can be used on survey design in general. Those recommendations range from detecting the users phone type, disabling or limiting the Flash design in survey and all the way to limiting the length of the survey question to avoid the mobiles phones with narrow screens.

They have also concluded, that smartphone market is going to grow and those smartphone users are also expecting shorter surveys. So they have made some more radical recommendations:

  • "Utilize larger sampling frameworks, while displaying shorter question sets to each respondent.  This process may entail presenting questions or question blocks randomly or in rotation to a subset of respondents.
  • Leverage prepopulation as much as possible.  Any fields that can be prepopulated eliminate the requirement to display another screen to respondents, thereby shortening the survey.
  • For concept testing, two to three concepts are more than likely the maximum quantity that will yield a respectable response rate in any one survey, in combination with the larger sampling framework that has already been described.
  • Trim project objectives and conduct more projects, rather than attempting to apply multiple objectives to a single project.  The data set may need to be married across projects on the basis of behavioral or other sampling methodologies, and not tied to individual respondents."

You can read more about the topic in their article, by clicking the link above to reach their blog article.

Short statistics of the October articles on blogs in our database in general:

  • from 115 blogs in our database, there were 61 blogs with one or more blog articles posted in October 2012.
  • The biggest count was 181 articles in October, posted on MRweb blog.
  • There were 94 relevant posts about web-survey methodology; and from those 94 posts, there were 21 quality articles in terms of web-survey methodology.

NOTE: We have discovered, that there are some blogs, which are not active at all and some of them even got deleted or they URL redirects to another blog (Zoomerang redirects to the SurveyMonkey for example). Some of those blogs have been temporarily disabled in our database and are only seen to the administrators. For the next edition of the "Blog of the month (November)", there will probably be less than 115 analyzed blogs in our database. We will inform you more about this topic in coming weeks.

Links:

Author: Marko Širec

This article has been read 2655 times

Comments