Observations on telephone survey HAC

Here are some observations in relation to the recent telephone survey on Hundertwasser Arts Centre

As expected given the research methodology, people under 40 are significantly under-represented in the survey sample – comprising just 10% of the unweighted sample relative to what appears to be 30% of the population.  A sample of n=100 carries a maximum margin of error of +/- 9.78%.  This means that reported opposition to the Hundertwasser Arts Centre (which I’ve calculated to be 43% in the <40 age group) could be expected to measure as low as 33% or as high as 53% if the survey were repeated 100 times.  So, the data is not very reliable at this level.  The only age group with a sizeable sample is the 60+ group – that most opposed to the Centre – so, really the only thing you can say with confidence from this research is that older people don’t support the Centre.


If you refer to the tables at the end of report you’ll also see just how skewed the sample is – there were just 31 people aged less than 30 years in the sample, compared with 540 people aged 60 years or more!


While the researchers have weighted the sample to represent the population as best they can, I would still have concerns about the extent of weighting required in relation to the youngest age groups (18-29 years and 30-39 years).  Weighting 3 times off a very low base can have the effect of amplifying errors in the original (unweighted) sample if this sample is atypical in any respect.


Some aspects of the survey design are also questionable.  For example: –


·         While people are (quite rightly) asked how much they feel they know about the Hundertwasser Arts Centre, it seems EVERYONE is asked whether they oppose or support it (including the 41% of the sample who claim to have no or little knowledge of what’s proposed) and EVERYONE is asked whether they support or oppose Council funds being used.  This makes no sense at all – how can someone who knows nothing about the Centre be expected to voice an opinion on it (certainly without further information being given – which would be normal practice).  Interestingly, there also doesn’t appear to be any analysis of claimed support/opposition to the Centre by extent of knowledge (that might show, for example, that people who know more about the Centre are also more supportive of it…). 

·         The Likert scales (Questions 5 and 9) both lead with the opposing view (Strongly oppose, oppose etc.).  While this isn’t incorrect as such – the researcher has to decide whether to start with the positive or negative position – this is probably less typical than to begin with the positive view, particularly when reading out response options as is necessary in a telephone interview (i.e. strongly agree, agree, etc.).  Interestingly, the question itself uses the more common order in saying “…whether you support or oppose the development…”.

Back to top