PLANNING FOR DATA ANALYSIS

The next aspect of the data analysis plan is to scan the data gathering instruments in the context of data analysis. During the process of selecting a method and technique and developing an instrument, the researcher is usually so oriented to content and procedure that he ignores the relatively minor issues in instrument development, such as whether or not their is a place for the respondent to indicate his name, age, class, or any other data that will be required for data analysis. After the data are collected and about to be analyzed it is often too late to capture this information. Thus, it is at this point of planning the data analysis that the researcher can profitability refer to the list below of characteristics of the data gathering instrument that will expedite data analysis.
1. There should be a place for name, school, class, age or any other classifying information the researcher will need in data analysis.
2. Every item should be numbered, and every option within an item should be separately numbered or lettered, for case of data analysis.
3. Options should not overlap on structured questions, that is, if one option is 5-10, the next should begin with 11.
4. It is preferable to have options circled rather than checked. This expedites the analysis of data and also eliminates the possibility that the check below means ‘high school’ to some respondents rather than ‘College’ as the researcher intended.
5. The order in which items appear on an instrument reflect the priority of information sought.
The researcher must recognize that respondents and observers tire and if they do the items which appear late in the instrument are answered less carefully or omitted entirely.
The order of the items has been planned with the processing of data in mind. If two pieces of data are to be analyzed simultaneously, this analysis can be expedited considerably if they are physically close together on the instrument. If some sub-analysis within the instrument is planned, such as computing the number of items correct on a sub-test or the number of positive responses on an interest inventory, the layout of the instrument should consider the placement of those items on each sub-test. 
The instrument should provide the respondent with a way of indicating inability or unwillingness to answer a question, record an observation, or respond to a measure. The possibility of pre-coding instruments, or printing them on different colour paper as an aid to data analysis should be considered. Instrument used in the pre-and-post administration of any technique should be easily distinguishable and so impossible to confuse.

Comments