The Ten Most Common Mistakes Made by Beginning Researchers

Project Size:
Attempting to answer a research question that is either too broad (or insufficiently precise). The first question has to be ‘what can I realistically achieve with the time and budget available to me?’. The second, ‘what do I really want to know?’.
Project Scope:
Embarking on research projects that involve questions already substantially answered elsewhere. Generating a focused research question is only the first step in scoping out your project. The second involves asking ‘what is already known about this question?’.
Being Insufficiently Objective:
Once a worthwhile research question has been agreed on, a common problem is attempting to reinforce the researcher’s preexisting opinion about that question. This gives rise to research designs that (usually inadvertently) attempt to validate (or refute) a particular perspective rather than waiting to see what the data say about the question at hand.
Mismatching the Method to the Question:
Research always need to be tailored made to the research question at hand. It makes no sense to select a methodological approach for reasons of personal preference or convenience.
Mismatching the Sample to the Method:
Research designs should be internally consistent process that enable researchers to construct a rigorous argument. It makes no sense to choose a particular research design but then use a different internal logic for choosing the sample (for instance, a survey design with a non representative availability sample).
Asking Poor, or Poorly Considered, Questions:
Leading questions, imprecise questions, and questions with poorly considered answer options are common mistakes in research. Question design needs to be a process, one whose outcomes need to be trialled before adopted wholesale (i.e., pretesting).
Prejudicing Analysis:
Also known as ‘cherry picking’, this involves the deliberate selection of those data from the results which support a particular point of view (see point #3) while ignoring those that do not.
Mismatching the Analytical Tools to the Data:
As for sampling, so for analysis. The techniques used to analyse the data need to match the kinds of data collected. It makes little sense to complete regression analysis to three decimal places with data collected in qualitative research using non-probability samples.
Making Analytical and/or Conceptual Leaps:
A common mistake in research is to asks about one thing but to conclude about another (most often seen where the research collects attitudinal data but concludes about concrete behaviours).
Taking the Research Results Out of Context:
All research is flawed and incomplete to some extent. These limitations need to be explicitly acknowledged and kept firmly in mind when talking about what your research actually shows.

Research First, Level 1 / 320 Manchester Street, PO Box 5596, Papanui, Christchurch 8542
Phone: (508) 4RESEARCH (508) 473 732 info@researchfirst.co.nz