Rural 5 - Evaluation Flashcards
Imagine you have some results, some show an increased change, some show a decreased, are they reliable?
Probably not, but it depends on the question…
What’s wrong with asking about people’s perception?
It may not accurately reflect reality
What’s wrong with giving people options on a survey?
It might not have included all the options someone might have included if they were given a free choice
What’s wrong with asking randomly selected residents?
You might not get a sample that accurately reflects the population, e.g. wrong distribution of age / gender / ethnicity
What’s helpful about a randomly sampled sample?
It’s unlikely you’ve asked the same person twice
What’s the problem with going on a ‘Wednesday’?
Only certain type of people will be present in your survey location, e.g. more elderly because families are at work/school
Identify problems with using bi-polar scoring systems?
Relies on subjective judgements, not a pre-calibrated scores. Hard to compare scores.
Judgements are not weighted on their importance (e.g. some types of deprivation are worse than others)
Descriptors mean different things to different people, so can be unreliable.l
What’s wrong with using random sampling?
You unintentionally introduce bias
Might be drawn to a particular social group - and over sample them - unreliable results
What’s wrong with systematic sampling?
You might miss groups of people - you’re sampling every nth interval - some views might not be recorded, could skew the results
What’s wrong with stratified sampling?
Need to access background population information (e.g. census) to identify the groups you want to sample
Don’t want to over-sample a particular group / want to remove bias.