Expert Evaluation Of The Research Methods To Study The Question-answering Process

By Author: Bunny
Total Articles: 11
Comment this article

The expert eveluation method include the behavioural coding , expert analysis and cognitive forms appraisal methods.
In market Research , Behaviural coding Involves the observers/coders coding interaction between the interviewers and respondents
during a Market Research interview. This procedure is aimed at identifying the questions that are difficult to ask or understand, lack of common understanding of terms and concepts and difficulties in processing the requested information.

Specific behaviours believed to reflect the problems that the respondents have with questions are observed. The coders specifically observe the question-answering process either as it occurs or review audio or video recordings of the interview.The technique allows a researcher to assign quantitative scores to question in terms of the numbers of problems coded for a question and the percentage of respondents experiencing each problem.

Informal observation is a variation of the informal coding in which the observer notes aspects of the respondent behaviour that seem interesting or unusual to the observer.

In expert ...
... analysis a market researcher who have not constructed the constructed and formulated the questionaire to be asked reviews a
questionaire to gather the understanding of the response task and to note potential problems . The questionnaires characterstics are described including the procedures for moving through the questionnaire , type of includign procedures for moving through the questionaire, type of respondent and nature of the tasks required for the respondents. List of problems that respondent might have and the types of mistakes they might make are made. The market researcher then classifies theb observations o get a general understanding of the questionnaire and the specific points where the difficulties may occur.

A number of factors enter the decision to select one mehtod of data collection over other.Budgets , speed of implementation , resources and staff of the data collection organization, sample coverage , expected response rates, length of the qustionnaire , types of question asked (open vs closed , sensitive vs non sensitive , complex vs simple ) , need for visual aids , expected social desirablity bias , expected risk of contamination of risks and interviewer effects are factors that enter the complex decision of selecting a method of data collection appropriate for a study. The benefits and limitations of each method ought to be weighed. This chapter has sought to review the various methods of data collection and their associated measurement errors. Despite the ever increasing knowledge of errors in surveys , it is apparent that much more has yet to be learned. In fact, while technologies advances have in some ways made the data collection task easier , these advances have opened up new areas for research on measurement error.

One Way to estimate measurement error directly is to compare the result of a survey measurement to another indicator of the same characterstic, one likely to be more accurate. There are two methods that are often used to do this ; one obtains external data on individual persons in a day.
another compares population parameter values with survey based estimates of the same quantity.

One conclusion from the diverse concepts of measurement error discussed above is the observation that notions of bias and variable errors take their meaning from uses of the data. To trivalize the point , reported years of formal education has one set of properties as an indicator of experience with formal schooling , and another set of properties as an indicator of cognitive abilities. Or from another perspective, the indentification number for a data set record can be considered to be a measure of income, one with very deep measurement error problems.
Meltdata Business Solutions. India