Recently, the first organizer of this session and Daniel Oberski received the Warren J. Mitofsky Innovators Award for the Survey Quality Predictor SQP 2.0 and its contribution to the improving questionnaire design. Although we appreciate very much this award, we think that there is still a lot of work to be done so that this tool can really play the role we hope it can play in survey research with respect to design of questionnaires and to correct for measurement errors in survey analysis.
In order to go further with this work, we invite all interested people, who have studied the SQP approach and would like to comment on its design or have applied the program and like to share their experience with others, to provide an abstract of a paper about their work for our session at the ESRA conference 2015. Papers about all different steps that are required to achieve the SQP program are welcome: MTMM estimation, meta-analysis of quality estimates, program itself and applications to improve surveys or correct for measurement errors…
1. Appling SQP 2.0 for researching the structure of basic values.
Mr Laur Lilleoja
Given study will try to analyses if correcting for measurement error changes somehow the structure of basic values. SQP 2.0 is used to get quality estimates for the value item, which makes possible to correct the model for measurement error and evaluate if it has an effect on structural level.
This study should give a general indication if correction for measurement error changes the structure of basic values or not. In the same time it also demonstrates how to use correction for measurement error in case of complex models, and additionally raises some potential problems concerning the correction procedure.
2. An application of SQP 2.0 to evaluate a self-control scale with data from the Cali Juvenile Delinquency Study, 2010.
Dr Alfonso Serrano-maillo
(Universidad Nacional de Educación a Distancia)
There is strong evidence that self-control is a robust predictor of delinquent acts. Grasmick et al.’s scale is the most usual way to measure this construct in Criminology. Some hypothesis regarding the elements and dimensionality of this scale are tested in this presentation. Analysis use both original data and data controlling for measurement error using Saris’ SQP 2.0. Hypothesis include that self-control is a second order unidimensional construct, as well as that findings is not an artifact due to measurement error. The use of SQP 2.0 includes making certain important choices, that will be discussed.