W. Dijkstra

Measuring the Agreement between Sequences

The present article proposes a new method to assess distances between sequences of states, belonging to, for instance, event histories. It is based on the number of moves needed to turn one sequence into another sequence. This number of moves is standardized by relating it to the number of moves needed to turn the sequence into its reverse. Weighting it in different manners by the number of common elements in both sequences yields three different agreement measures, suited for different kinds of research questions. Life history data from 494 young adults show that applying our agreement measure to these data can yield insight in interesting relationships, otherwise not readily obtained.

Interviewing Style and Respondent Behavior

Guided by the model of the question-answer process as developed by Cannell and his coworkers, our main research question focused on how interviewers should behave in order to obtain adequate responses in survey interviews. A field experiment was conducted in which the interviewers adopted one of two behavioral styles, namely, a personal versus a formal style of interviewing. The dependent variables concerned measures of accuracy, social desirability, and the degree of relevant as well as irrelevant information. In addition, by experimentally manipulating the interviewers alleged opinion, it was possible to ascertain conformity effects. Two different hypotheses were proposed to account for the effect of interviewing style on the dependent variables. The motivation hypothesis proposed that a personal style, compared with a formal style, would motivate the respondent to do well on the task and to give responses that would be as accurate and as adequate as possible. According to the ingratiation hypothesis, respondents interviewed in a personal style would be more inclined to attempt to ingratiate themselves with the interviewer, leading to more socially desirable responses, conformity, and irrelevant information. The results supported the motivation hypothesis. We contend that both motivation and ingratiation are important mechanisms in the question-answer process. However, it may be easier to influence the respondents motivational level by means of the style of interviewing or other interviewing techniques than the respondents ingratiation tendencies.

Styles of Interviewing and the Social Context of the Survey-Interview

Of all data collection procedures in the social sciences, the survey-interview is most frequently used (cf. Brown & Glimartin, 1969; Wahlke, 1979). Considerable doubt has been expressed, however, concerning the validity and reliability of the information it yields (e.g., Phillips, 1971). The term “response effects” refers to the effects of variables that influence or distort the responses, such as the race of the Interviewer, the social desirability of the response alternatives provided, or the way questions are formulated.

Types of Inadequate Interviewer Behaviour in Survey-Interviews

Researchers have long been aware that the information gathered by means of survey-interviews is probably distorted. Perhaps Rice (1929) was the first who was able to show, as early as 1929, that interviewers can affect the respondent’s answers in a definite way. Since that time the number of studies on information-distortion in interviews has constantly increased (see for example Sudman and Bradburn, 1974; Dijkstra and van der Zouwen, 1982). By now there is clear evidence that the respondent’s answers can be the result of numerous factors other than the respondent’s opinion, attitude or whatever the investigator is interested in.

Linked Markov Sources Modeling Outcome-Dependent Social Processes

Many social processes are adaptive in the sense that the process changes as a result of previous outcomes. Data on such processes may come in the form of categorical time series. First, the authors propose a class of Markov Source models that embody such adaptation. Second, the authors discuss new methods to evaluate the fit of such models. Third, the authors apply these models and methods to data on a social process that is a preeminent example of an adaptive process: (encoded) conversation as arises in structured interviews.

How to Measure the Agreement between Sequences A Comment

Some problems of optimal alignment procedures to measure the agreement between sequences are discussed. Hidden Markov models may be a new approach that is especially suited for grouping sequences that are more or less alike.