Your data analysis is only ever as good as the thought and methodology behind data collection. If you're collecting the wrong data or the right data for the wrong reason, your analysis is useless at best.
This morning the More or Less program on BBC Radio 4 also got into it about a public consultation the government did this summer, surveying business owners (including me). The government wants to go back to imperial measures. They need the outcome of the consultation to support their desire.
Some of the questions in the consultation were stunningly biased. I felt steamed about it at the time. Asking whether you would prefer things to be in imperial units or in imperial units alongside (less prominent) metric units, with no other options, was one example. More or Less brought in an expert on survey design and asked questions as though clueless (which they definitely aren't, data is their thing in life). Cool, not inflammatory at all, but very revealing. Bad data = worthless consultation. Exactly as you said.
There's always a balance between what can be measured and what can be observed or intuited. (Reminds me of Special Agent Gibbs, NCIS TV show and his famous "gut instinct.") I'm a fan of Temple Grandin - I heard her years ago on NPS (I think it was a Terry Gross interview) and saw the movie about her. I don't know what you mean by "affiliate link" though. She is an example of how a "disability" is a gift She also learned, from studying how to keep cattle less stressed as they are being led to slaughter, how to soothe herself with a squeeze box. Very enlightening.
I remember Grandin's explanation about the squeeze box. The way she thinks is fascinating.
A little girl next door to us is autistic. Her mother is in awe of her and says autism is her superpower.
An affiliate link means that if you use that link to buy one of her books, I will get a commission. Pennies, probably, but it's a financial interest so I declare it.
Lies, damned lies, and statistics.
Your data analysis is only ever as good as the thought and methodology behind data collection. If you're collecting the wrong data or the right data for the wrong reason, your analysis is useless at best.
This morning the More or Less program on BBC Radio 4 also got into it about a public consultation the government did this summer, surveying business owners (including me). The government wants to go back to imperial measures. They need the outcome of the consultation to support their desire.
Some of the questions in the consultation were stunningly biased. I felt steamed about it at the time. Asking whether you would prefer things to be in imperial units or in imperial units alongside (less prominent) metric units, with no other options, was one example. More or Less brought in an expert on survey design and asked questions as though clueless (which they definitely aren't, data is their thing in life). Cool, not inflammatory at all, but very revealing. Bad data = worthless consultation. Exactly as you said.
There's always a balance between what can be measured and what can be observed or intuited. (Reminds me of Special Agent Gibbs, NCIS TV show and his famous "gut instinct.") I'm a fan of Temple Grandin - I heard her years ago on NPS (I think it was a Terry Gross interview) and saw the movie about her. I don't know what you mean by "affiliate link" though. She is an example of how a "disability" is a gift She also learned, from studying how to keep cattle less stressed as they are being led to slaughter, how to soothe herself with a squeeze box. Very enlightening.
I remember Grandin's explanation about the squeeze box. The way she thinks is fascinating.
A little girl next door to us is autistic. Her mother is in awe of her and says autism is her superpower.
An affiliate link means that if you use that link to buy one of her books, I will get a commission. Pennies, probably, but it's a financial interest so I declare it.