Help required with estimating sample size for online poetry questionnaire

Hi all, I am writing to you because I need advice regarding a poetry questionnaire that you can find here: Imagery questionnaire (google.com).

As I am now collecting responses, it dawned on me that I don’t know if there is a formula to estimate sample size for questionnaires. I have found some papers online (including the one you recommend in another thread of Concordance) but they are very technical and I don’t know if they apply to my case. So basically this is my design in a nutshell:

  1. materials: a set of 8 poems; each participant reads 2 poems randomly chosen by the program and answers some questions

  2. for each poem a mixture of tasks is required, mostly open questions and Likert scales.

  3. the study is exploratory, the stimuli materials were not chosen according to specific hypotheses (e.g., low vs high imagery poems) but rather to represent the variety of the ‘poems’ population.

  4. I am interested in poetry reading in general (so the set of 8 poems) but also on each specific poem (to account for differences in style). So I am interested in an item-analysis, not in participant-analysis. I want to see how responses to each poem compare, not how specific individuals or groups of individuals respond to poems.

For now I have 73 responses but I think I need more: so far, each poem has been read, rated and commented by an average of 18 participants. Is it enough? are there any recommended standards?

A related question concerns the recruitment of participants in ways other than word-of-mouth and online postings: I have seen that some studies use either Amazon Mechanical Turk or Crowdflower, but I guess funding is needed in these cases, and I don’t have it, unfortunately. What would you recommend doing?

Thank you in advance for your help!
Davide

Hi David,

yes, there are formulae to calculate how many recipients you need, but I am rather ignorant of them. But I would say: just start!

Would not a call for help on the PALA website be of help???

Willie

Hey, Davide.
I am using GPower 3.1. to calculate parameters depending on statistics I am going to apply.
You can check it here:
https://www.psychologie.hhu.de/arbeitsgruppen/allgemeine-psychologie-und-arbeitspsychologie/gpower

I hope I may be helpful in some aspects.

Best wishes,
Danil