SCV survey, processed by Statistics Flanders
In determining the general level of cultural participation, 10 activities are involved: attending or visiting concerts (4 separate genres), museums, libraries, cinemas, dance or ballet shows, theatrical and operatic performances.
The ‘core audience’ participates at least 3 times per year in 3 of the 10 activities.
‘Interested participants’ take part at least once a year in 3 different types of activities.
‘Incidental cultural participants’ partake in culture to a less intensive and less varied extent.
The ‘regular cultural participant’: the sum total of the ‘core audience’ and the ‘interested participants’.
For the long time series for specific cultural activities, in all cases a distinction was only made between participants and non-participants. It was impossible to obtain greater detailing in a similar manner for every survey year. But even with this rough distinction, there are differences in the survey questions which may have an impact on the answers returned.
Cinema visits: for this activity, the questions have remained fairly stable over the years, producing fairly comparable time series.
Museum visits: the question was broadened over the years, involving a break-down into Belgian and foreign museums and the addition of ‘or a gallery’ in the wording of the question. In spite of this broadening, the rough break-down into participants and non-participants is fairly comparable over the years.
Library visits: no figures for 1996 and 1999. For the other years, changes were made to the answer categories, although the rough, general level of participation is fairly comparable over time.
Concert visits: major differences in the way the question has been worded over the years. Since 2011, a prior filter question has been asked before enquiring into the respondents’ attendance of specific concerts. Before 2011 (from 2000 onwards), no filter question was used and the level of participation was established based on the attendance of 5 concrete types of concerts (classical, pop-rock, jazz-blues, traditional-folk and opera). For the 1996-1998 time frame, the level of participation had to be established based on 2 questions enquiring into the respondents’ participation in / attendance of a “pop concert” and “musical concert or opera”. In 1999, the level of participation was determined based on 5 questions, but these differed from the questions that were routinely included as standard from 2000 on.
In the case of concert visits, there are clearly strong query effects. This was also demonstrated in an experiment conducted in 2009 and 2010 whereby half of the respondents were presented with the old questions and the other half was asked to answer the new questions. All of which means that the comparability for this particular variable is not quite good. The changes in the wording of the questions also clearly show in the graph.
Performing arts: this question too has been broadened over the years, involving a break-down into amateur and professional troupes, and theatre and dance. In the early years, there was only 1 question pertaining to the attendance of “a theatre play or a ballet performance”. Despite this broadening, the general degree of participation is fairly comparable over time. Nonetheless, it cannot be ruled out that the broader questioning also affects the answers returned.
Remarks on quality
In order to be able to compare the longer time series of the specific cultural activities, the population was restricted to Flemings aged 18 to 75 who have the Belgian nationality. For the general cultural participation, however, the population covers all over-18s (including non-Belgians).
The SCV survey ‘Sociaal-culturele Verschuivingen in Vlaanderen’ (Socio-Cultural Shifts in Flanders) is an annual face-to-face survey conducted by way of a random sample survey among Dutch-speaking residents living in the Flemish Region and in the Brussels Capital Region. The survey enquires into the opinions, beliefs and behaviours with regard to social and policy-relevant topics. Each year, some 1,500 exploitable interviews are taken. The response rate is around 60%, but this figure fell to 55% and 50% in the survey years 2017 and 2018 respectively.
Quality guidelines are assiduously respected, both in drawing up the questionnaire, the sampling, the data weighing and in interviewer training. It is important for the user to know that the percentages based on survey data are estimates of the corresponding population percentages within a given interval. The size of this interval depends on the percentage obtained and on the sample size. For a sample size of 1,500 respondents, the reliability interval for a reported percentage of 50% is approximately 5%, i.e. between 47.5% and 52.5%. For reported percentages of 20% or 80%, this interval is narrower, but for smaller groups (e.g. one level of education or one household type) the interval is larger.
All percentages and averages are weighted. The weights used in this respect are a compensation for the over- or under-representation of certain categories. This over- or under-representation may be the result of sampling errors or differences in non-response, among other things.
Statistics Flanders: SCV-survey