Different methodologies for party positioning and pay attention to the evaluative criteria of validity, reliability, verifiability, costs and utility for comparative research
Expert Survey
Reliability: can be measured by amongst others, calculating the standard deviation over the estimates of all experts. A large standard deviation is an indication of unreliable measurement.
Validity: Assessing the validity of a party position measure is more problematic since we lack an
uncontested standard against which to compare the expert opinion
Costs: The expert survey is very cost-effective (Marks et al. 2007). Experts are able to integrate large amounts of information from multiple sources and signal the emergence of new issues and estimate their importance
Verifiability: the expert survey enables researchers to use the aggregated knowledge of several (country) experts. The downside of this feature is, however, that it is impossible to trace the sources on which the calibrations of the parties are based since the experts are usually not asked to list the information or knowledge they have used to position the parties. This constrains the researcher, and third parties, who cannot verify the coding.
Comparative Research: the comparability of expert surveys is difficult because the surveys often contain different questions and different formulations of similar questions, which makes it hard to integrate them into one dataset
Other: the source of the positional estimate remains unclear and is likely to vary considerably among experts. Methods of cross-validation of concepts and knowledge levels can be incorporated in expert surveying, improving their validity and reliability. Yet, cross-national and cross-time expert surveys should be used with caution as the lack of transparency of sources for the calibrations turns the expert survey into a black box.
Voters
3 Methods for Asking Voters to Position Parties: 1) asked about their party affiliation and then to position themselves on an issue or dimension, or 2) they can be asked to place the political parties on these issues or dimensions. 3) A better, third, method is to ask voters to both place themselves and all the parties of the political system.
Sources of Bias and problems in eliability and validity:
The most important problem with mass surveys is the lack of control over interpretation of concepts in the questions, as respondents are likely to project their own preferences and interpretations on the questionnaire
An additional source of bias results from the varying levels of participation by groups of voters with different levels of political knowledge and interest
An additional source of bias results from the varying levels of participation by groups of voters with different levels of political knowledge and interest
Problems also arise when conducting cross-countries surveys
elite Survey
click to edit
There is an obvious advantage of asking elites to
position themselves. But it is problematic:
For one, it assumes that parties are unitary actors. Parties may be internally divided and factionalised, so it matters whom inside the party is asked to provide the ‘official’ position
party representatives at different positions may take different considerations into account when calibrating their party
the inner-party power relations vary substantially across countries