KRAZY KATS

Welcome to Krazy Kats - a friendly informal online community discussing life issues that we care about. Open 24/7 for chat & chill. Come and join us!

2 posters

    Stuff yer Statistical Analyses!

    Kitkat
    Kitkat

    Stuff yer Statistical Analyses! Empty Stuff yer Statistical Analyses!

    Post by Kitkat Tue 04 Dec 2018, 14:23

    Makes me cringe, sometimes, the sort of ideas and 'results' that are spread around the media these days - so-called 'surveys' and fill-in vacuous articles that are concocted and planted into people's heads and presented as facts.

    An example here, from today's BBC News:

    'Most adults living unhealthy lifestyles'
    https://www.bbc.co.uk/news/health-46439892

    The article starts off with "The data from the Health Survey for England showed nearly nine in 10 had at least one unhealthy trait" ...

    "nearly nine in 10" Huh? How do they work that one out? - What exactly does nearly nine in 10 people mean? shrug Nearly nine can only mean 8 where people are concerned. The only other consideration there would be perhaps eight and a half, eight and three-quarters, or pick any other fraction of your choice ... So, the screaming question there is - How on earth do you define eight and two-thirds (etc) of a person??

    Certainly has me confused. surprised Another question I have, is - Where has this 'statistical' information come from? What source have they used to glean such details? What section of humanity have the 10 (or nearly 10) individuals been picked from? In this particular case, the Health Survey for England - How many comparable groups of 'nearly 10' people all around England have been selected to comprise the basis of such a study?

    Am I being too analytical? Is that not what 'statistics' are (or should be) all about?
    lar-lar
    lar-lar

    Stuff yer Statistical Analyses! Empty Re: Stuff yer Statistical Analyses!

    Post by lar-lar Tue 01 Jan 2019, 22:27

    The gap between nearly 9 and 9 can be significant. It's not unusual to see this sort of thing.

    You might like this:

    https://www.investopedia.com/terms/r/representative-sample.asp


    'How many comparable groups of 'nearly 10' people all around England have been selected to comprise the basis of such a study? Good question!

    It's interesting many many studies are done on university students..hardly representative..think about the age range!

      Current date/time is Thu 28 Mar 2024, 18:54