Timeline for Margin of error for a specific question on a survey?
Current License: CC BY-SA 4.0
18 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 9, 2019 at 20:30 | comment | added | user233540 | That's kind of mind blowing, even if we sent a survey to 100 folk, and 30 respond, we can't treat those 100 as a sample in itself? We can't derive a margin of error from it? In other words, make inferences from the people who got sent surveys with those who actually responded back | |
| Jan 9, 2019 at 20:22 | comment | added | StatsStudent | That being said, in general you the margin of error for a proportion when sampling IS INVOLVED (this is NOT what you have) can be found here: newonlinecourses.science.psu.edu/stat414/node/264 assuming a simple random sample. | |
| Jan 9, 2019 at 20:20 | comment | added | StatsStudent | There is no margin of error. That term only applies when there is sampling. In your case, you'd want to use the standard deviation that I provided in the second comment. | |
| Jan 9, 2019 at 20:19 | vote | accept | user233540 | ||
| Jan 9, 2019 at 20:18 | comment | added | user233540 | Yes, you've been very helpful. Just curious, what is the margin of error formula for my case above in which case I have no sample? do I use the same formula but instead of sample, I use my actual population size? | |
| Jan 9, 2019 at 20:09 | comment | added | StatsStudent | Even though only 15% respond, you are still not sampling. Sampling involves the mechanism by which you allow individuals to be eligible to respond - not the mechanism that actually creates the response. In addition, if you are simply estimating the percentage of those who would recommend (and not totals), without having "auxiliary information" on the non-respondents as well as the respondents(e.g. demographic information or purchase patterns), then you wouldn't be able to even revise your original calculation of the percentage of people who would recommend. Does that make sense? | |
| Jan 9, 2019 at 20:00 | comment | added | user233540 | Ok thank you, very helpful. but I do want to clarify, not everybody who we send surveys to responds. We only have a 15% response rate, so we are sampling no? | |
| Jan 9, 2019 at 19:57 | comment | added | StatsStudent | The 400 are the number of respondents -- those who chose to respond. There's no sampling error associated with those people since everyone was sampled (e.g. 100% sampling). If you can assume that your non-respondents are similar to your respondents then your job is done. You can calculate the proportion of people likely to recommend the product to friends and family and that is your population value -- there is no uncertainly about it - so there is no margin of error. | |
| Jan 9, 2019 at 19:48 | comment | added | user233540 | So just a little confused, then what is the 400? How would I calculate the margin of error that the 400 represents for the population? | |
| Jan 9, 2019 at 19:36 | comment | added | StatsStudent | @user233540. Well in that case, the population would be 1,000 and the sample size would be 1,000. So this wouldn't be a SAMPLE survey, but a complete CENSUS. But only 400 responded. | |
| Jan 9, 2019 at 19:33 | comment | added | user233540 | Can I ask one more clarifying question? If I sent out a survey to 1,000 users of Product A, just once, not on a rolling basis, to get their feel for product A. And we get 400 results back, can I say the population is 1,000 and the sample size is 400? | |
| Jan 9, 2019 at 18:40 | history | edited | StatsStudent | CC BY-SA 4.0 |
added 4 characters in body
|
| Jan 9, 2019 at 18:34 | comment | added | StatsStudent | You can use the results, but there is nothing to sample: you've asked every single person the question. What I would simply do is simply calculate the average response of the responses you received so far and calculate a population standard deviation= $\sum_{i=1}^{200}(x_i-\bar{x})^2/200$, where $\bar{x}$ is your mean response to Q10, to get am understanding of how much your responses vary. If you think the 200 responding are similar to the 100 who didn't respond, then that IS your population statistic and hence there is no uncertainty around it's estimation. | |
| Jan 9, 2019 at 18:29 | comment | added | user233540 | The product can't be purchased multiple times. At least not in a long time period, its software with 5 to 10 year licences. | |
| Jan 9, 2019 at 18:11 | history | edited | StatsStudent | CC BY-SA 4.0 |
add final paragraph.
|
| Jan 9, 2019 at 18:10 | comment | added | user233540 | So the last questions is actually: "Based on your experience implementing this product, how likely are you to recommend our product to friends and family?". So its a Net Promoter Score question, answers can range from 0 to 10. Let's ignore the 700+ people then, can I make an inference from the results I've gotten back so far? Can I use the 300 results I have gotten back over the last 6 months? Or am I still not able to sample? | |
| Jan 9, 2019 at 17:58 | history | edited | StatsStudent | CC BY-SA 4.0 |
added final non-respondent example
|
| Jan 9, 2019 at 17:50 | history | answered | StatsStudent | CC BY-SA 4.0 |