Hi all,
In my survey, I have two questions where I am unsure about the type of measurement (ordinal or scale).
- My first question is: 'On average, how many days a week did you work from home?'.
Answer options are: 0 days, 1-2 days, 3-4 days, more than 4 days
- Next question: 'How much time do you spend per day on video conferencing during telework?'.
Answer options are: Less than 1 hour, 1-2 hours, 2-3 hours, 3-4 hours, more than 4 hours
Could you please tell me which measurement should I use in SPSS for these two answer options?
Thanks in advance!
Hello Dilnas,
Both examples are defensibly ordinal strength (e.g., 1, 2, 3, 4) as they stand.
Up to a point, you could make these interval, by assigning the midpoint of each range as the scaled value. For example, "3-4 days" could be given a score of 3.5, whereas "1-2 hours" could be given a score of 1.5. "0" would be scored as 0.
This approach obviously can fail when you try to quantify the highest category, unfortunately, given that they are open-ended. For the days per week example, you could: (a) censor the values as all being "4;" (b) assign midpoint of 4-max (which would be 4-7, or 5.5); or (c) go back and interview 25-30 people who selected the "4" option and ask them, individually, how many days (more precisely). Use the median value of these responses as your scale value.
I suppose this is a good example of why it makes sense to have measurement specialists review an instrument and then do a small trial run before going live with a survey.
Good luck with your work.
It would be worth knowing why you didn't ask for respondent to give the number of days in the previous week (which using Mosteller and Tukey's framework is a counted fraction, and the suggest the logit) and how much time a time (which might be a count) rather than forcing respondents into those categories. The choice of categories affects how people respond and can also affect responses to other questions (Article How response alternatives affect different kinds of behaviou...
). I guess the main question is why do you want to know the "type of measurement"? Before answering it would be worth consulting:Article Nominal, Ordinal, Interval, and Ratio Typologies Are Misleading
The straightforward answer for your situation is “ordinal.“ There’s no need to guess the exact count or duration after you’ve collected the data.
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS