Over the past few weeks, I’ve read a few dissertations in their draft forms. They had several challenges in common.
Let’s examine five mistakes they made, and how to fix them.
1. The dissertations took several pages to describe the characteristics of their participants. Each demographic was described in detail with an accompanying table and or graph that said exactly the same things in duplicate or triplicate.
What’s wrong here? APA guidelines recommend that tables, graphs and commentary must be complementary, i.e. add information and avoid redundancy.
What to do? It is much more efficient to use a single, well-formatted table to describe all the demographics and characteristics of the study participants at once. One page should be sufficient.
2. Some dissertations gave a separate frequency table for every one of their 5-point Likert-type items. One document had an additional 33 pages of these detailed tables pasted straight from the output of the statistical software!
What’s wrong here? Too much detail and no summary.
What to do? Provide a single table or graph for all items intended to measure the same construct, with frequencies summarised as percentages. Then add succinct commentary to this overall table or graph.
3. The dissertations calculated Cronbach’s Coefficient alpha across all the Likert-type items in the questionnaire.
What’s wrong here? Cronbach’s alpha is only meaningful if calculated using items designed to measure the same construct.
What to do? Calculate alpha only for items intended to measure a single construct.
4. The dissertations claimed that they were using a measurement instrument that was valid based on what others said about it in the published research literature.
What’s wrong here? Validity is not defined as a property of an instrument or test. Rather, validity is defined in terms of whether its scores are useful and relevant for the purposes of your research.
What to do? Discuss the validity of the test scores as used for your research, and also mention how the test scores have been used in published research.
5. The Abstracts were too long and poorly structured, comprising several paragraphs instead of one.
What’s wrong here? The abstract should not be longer than 250 words.
What to do? Write the abstract on its own page, in a single paragraph, under the heading “Abstract” which is centred. Include the research problem, the study method and research design, the sample size, primary analysis, main findings and implications.
Let me know if I can help you with your dissertation or thesis.
Contact [email protected]