Subscribe via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.


Fun with research

surveymarker***EDITED TO ADD: THANK YOU, THE DRAFT SURVEY IS NOW CLOSED! Stay tuned for more news on this project! ***

Do you like Undina’s posts about perfume stats of various sorts!? Of course you do! We’re perfume nerds and we love data about perfume.

As it happens some buddies and I are studying survey design and stats over the summer, and I would like to do a “real” exercise. What better to use as a sample research topic than perfume?

One of the steps we’re learning is to TRY OUT your questions first and see how they do with a small group. You should see where people abandoned answering, if you worded anything incorrectly, etc.

Well, there’s only a small group of you, my peeps, reading this blog – so let’s try it! I’ve drafted up a research survey (hopefully according to the guidelines of good research design). Please try it out, and feel free to post comments to me about it below. DO NOT SHARE at this time – if it works out well, I WILL ask you to share a future draft, and then we should be able to get some good data for analysis (if my stats homework pays off!).

The Unseen Censer’s Hopefully Scientific Survey DRAFT (should take 5 minutes or less)

And thanks for your assistance in my nerd-fun!

***EDITED TO ADD: THANK YOU, THE DRAFT SURVEY IS NOW CLOSED! Stay tuned for more news on this project! ***

Image is “Survey marker 135920” by melanie cook, via Flicker; used under Creative Commons license, some rights reserved.

Bookmark or Share

5 comments to Fun with research

  • I completed the survey; it was short and well presented. When you said DO NOT SHARE, you mean don’t share the topic of the survey, correct?

    • Judith

      Hello Hajusuuri! I meant don’t tweet it or blog it; I only want some sample answers to test (which has already been very helpful in terms of the exercise: putting into practice some of what I’ve learned about survey design). I’ve also realized one of the toughest lessons already: how to phrase in a politic way some of the results? I have to think harder!!
      I’m going to close this survey tomorrow and take a look at what I have and perhaps do some tweaking.

    • Judith

      PS THANK you! 🙂 And PPS: yes, not sharing the topic of the survey is also helpful. 🙂

  • Done.
    A couple of thoughts (but feel free to disregard them: I’m an amateur and do not have any systematic knowledge on the subject)

    Question 1: You completely disregard people who answer “No” to the first question. It’s both not nice (they’ve already decided to participate and clicked on the link!) and wasteful (you have a subject – why to let them go?). Maybe instead for those who answers “No” you could offer at least one additional question in the end. For example, something like: “Why do you think you had never had this experience?” With answers: 1) I do not wear perfume in public; 2) I wear only very inoffensive perfumes; 3) I apply perfumes very lightly; 4) Nobody dares 😉 etc.
    Also, you do not need to give the instruction “Choose one” with radio button where more than one choice isn’t technically possible.

    Question 2: You’re asking “check all that apply” but all the following questions are about the most recent experience. You might get more interesting data if you split the question into two separate – 1) how many times (“once”, “2-3 times”, “they do it all the time!!!”, etc. and 2) when was the most recent occurrence? (change it from check boxes to radio buttons and add an option for more than a year ago).

    With all other questions I’m not sure how you plan to analyze these answers: it seems you do not have enough qualitative variables. Also, it’s much harder both to get and to analyze free text responses. It might be easier for people to respond if you gave them more pre-set choices (with a possible “Other” and write-in). For example, for Question 4 I would ask: “If they said they did not like your perfume, what was the main reason?” 1) no reason, they were just being mean 2) it reminded them of somebody/something unpleasant 3) they claimed headache/asthma 4) other – and maybe a free text field to add comments not only for “other” but for any of the answers.

    I would probably add a couple more questions on the topic (again, easy to answer and to group answers) – the relationship to the person who disliked the perfume (spouse, parent, relative, friend, co-worker, stranger, other); whether it changed respondent’s feelings about that perfume and whether they stopped wearing it in presence of the “critic”?

    Once again, sorry if I’m overstepping.

    • Judith

      Hello dear!

      I’m not going to explain all my question design choices because I think some of that explanation would invalidate some of my results (and I’m sure the same people will be back here to take the “real” survey). Some of your points are very helpful though! I do think you’re right that I should do something with people nice enough to answer the survey at all, but your answers are 1) not on my topic and 2) very leading. 🙂

      Yes, I deliberately chose to put “Choose one” so that people would not be confused as I switch back and forth between questions where there is more than one answer possible and questions where there is not. I do not want to confuse or upset anyone nice enough to fill out my survey, and their job is to answer, not figure out how the survey questions work! I have a few open-ended questions, which I will have to code for analysis, as you say, but certainly don’t want more. I am just trying to address a small easy topic here. (For instance I’m not interested in the answers to any of your questions in your next to last paragraph – but it sounds like you are! Makes sense, of course.)

      It’s interesting that you say I don’t have enough qualitative variables; all the information is qualitative, and subjective. I’m going to see if I can successfully do a correlation analysis, and infer some conclusions (of which I’m sure you’ll point out the flaws).

      And while, as I say, I’m not going to explain every design choice, I absolutely do not want to imply that everyone has had the experience of someone disliking their perfume. In fact until last week I would say I’ve never had that experience, except from my hubby whom everyone knows dislikes perfume – which is why I had the idea for the survey. I see a lot of people talking on the web boards about others disliking their perfume but without much direct experience of it I have really no idea how often that occurs and why. So I thought I’d do it as my sample survey.

      Not overstepping at all! I’ll send you the references for books I’m studying if you’d like to see them. (I may actually footnote them in the final piece! We’ll see how ambitious I get.)

What do you think?