Gransnet forums

Coronavirus

How is COVID-19 affecting British opinions, jobs and wellbeing?

(13 Posts)
janipat Fri 27-Mar-20 23:10:15

An interesting survey GGMK3 thanks for sharing. I understand why people wonder at the accuracy of what appears to be a small sample, after all 3000 out of a population of over 67 million seems ridiculously small. As long as the selection of participants follows scientific guidelines it should be fairly accurate, but for anyone not trained in statistical analysis it can be very baffling. DH is such trained, I found it totally baffling before he explained, many, many years ago!

GracesGranMK3 Fri 27-Mar-20 22:54:22

I agree with your opinion of YouGov growstuff but it is still good to see there is nothing unexpected in their results.

Pikachu Fri 27-Mar-20 22:23:15

Yes, thank you.

growstuff Fri 27-Mar-20 22:22:33

Thank you for posting the link GracesGran. YouGov is not always the most "scientific" of polling sites, but I can't see anything surprising or disputable in the results of the poll.

Maybe SalsaQueen could explain why she disagrees with the findings.

GracesGranMK3 Fri 27-Mar-20 22:20:03

I do.

GracesGranMK3 Fri 27-Mar-20 22:19:51

I d hope you both benefited from the linked article.

GracesGranMK3 Fri 27-Mar-20 22:18:51

That's okay then.

Pikachu Fri 27-Mar-20 21:54:02

Actually 3,000 is quite an acceptable sample.

GracesGranMK3 Fri 27-Mar-20 21:50:29

As a generality 10% of the population you are testing, as long as it doesn't exceed 1,000 is an excepted size to get a meaningful result. If anything 3,000 is well over the size usually seen as acceptable - or affordable for any increase in detail you would get.

I wasn't "offended". I just get fed up by unhelpful negative comments on GN. Fed up isn't offended.

SalsaQueen Fri 27-Mar-20 21:38:28

GracesGranMK3 In the big scheme of things, 3000 isn't many.

I'm not knocking your post, merely pointing out a fact. If you're fed-up by that, you must be easily offended.

GracesGranMK3 Fri 27-Mar-20 18:53:26

3,000 is quite a large sample SalsaQueen. Where did you learn about statistics? It is quite reasonable to extrapolate from that number.

Do you know, I can't tell you how fed up I feel when I have put something up that people might, or might not, find interesting only to be met by the standard knocking post.

SalsaQueen Fri 27-Mar-20 18:41:21

The survey doesn't represent "many" people. Only 3000 people were asked!

GracesGranMK3 Fri 27-Mar-20 17:11:30

I thought some of you might be interested in some recent YouGov research here