PROCEDURES FOR CODING DATA

A coding frame is simply a set of instructions for transforming data into codes and for identifying the location of all the variable measured by the test or instrument.    Primary data gathered from subjects and informants is amenable to control during the data collection phase.   The implication is that highly structured data, usually derived from tests, questionnaires and interviews, is produced directly by means of a calibrated instrument or is readily produced from raw scores according to established rules and conventions. Generally, measures such as physical characteristics such as height and weight are measured on the ratio scale.    Whereas psychological attributes such as measures of attitude and standard dimensions of personality are often based on questions to which there is no appropriate response.   However, the sum of the responses is interpreted according to a set of rules and provides a numerical score on the interval scale but is often treated as though the measures relate the ratio scale.   Norms are available for standard tests of physical and psychological attributes to establish the meaning of individual scores in terms of those derived from the general population.   A questionnaire aimed at determining scores as a measure of a psychological attribute are said to be pre-coded; that is, the data reflects the coder’s prior structuring of the population.   The advantages of pre-coding are that it reduces time, cost and coding error in data handling.   Ideally, the pre-coding should be sufficiently robust and discriminating as to allow data processing by computer.

A coding frame should include information for the variable to be measured:

1. the source data;
2. a list of the codes;
3. column location of the variable on the coded matrix.

Thanks