Chapter 3. Problem Statement, Aim, Objectives and Questions
3.1 The problem statement
A problem statement explains the practical or scholarly issue, what evidence indicates that the issue matters, what is not adequately understood, why the uncertainty matters, and what the proposed study will do. It should not simply assert that a topic is important. Support key claims with credible, traceable sources and local evidence where appropriate.
A useful structure is: (1) context and importance; (2) documented problem; (3) what existing studies have established; (4) specific uncertainty or gap; (5) consequence of leaving the gap unresolved; and (6) the study’s intended response. The problem statement must not pre-judge the result.
3.2 Aim and objectives
The aim is the overall purpose, usually expressed in one sentence. Objectives divide that aim into achievable analytical tasks. Use verbs that can be evidenced, such as estimate, compare, examine, test, assess, determine the association, or evaluate predictive performance. Avoid objectives such as “understand everything about” or “prove that”.
Illustrative aim: To examine the association between perceived supervisor support and work engagement among full-time employees in selected service organisations.
Illustrative objectives: (1) describe the distribution of perceived supervisor support and work engagement; (2) assess the reliability of the selected scales in the study sample; (3) estimate the association between the two constructs; and (4) assess whether the association remains after adjustment for prespecified covariates.
3.3 Research questions and hypotheses
A research question states what the study will answer. A hypothesis is a testable prediction derived from theory or prior evidence. Not every quantitative study requires formal hypotheses: descriptive studies may be guided by questions, while explanatory and confirmatory studies often specify hypotheses. Hypotheses should be written before examining the results whenever possible.
| Question type | Example | Possible analysis |
|---|---|---|
| Descriptive | What is the level of engagement in the sampled workforce? | Frequencies, mean/median, dispersion, confidence interval |
| Group comparison | Does engagement differ by work arrangement? | t-test, ANOVA, or suitable non-parametric/model-based alternative |
| Association | Is support associated with engagement? | Correlation or regression |
| Prediction | How accurately do selected factors predict churn? | Regression with validation and performance metrics |
| Intervention | Does training improve performance compared with usual practice? | Randomised or quasi-experimental comparison, adjusted for design |
3.4 Alignment audit
Every objective should map to a question, variables, measure, sample requirement and analysis. If an objective has no planned analysis, remove or revise it. If an analysis does not answer an objective or question, justify it or remove it. This alignment table is one of the most useful tools for preventing a dissertation from becoming a collection of unrelated statistical outputs.
| Objective | Research question | Variables/measure | Planned analysis |
|---|---|---|---|
| Describe engagement | What is the engagement distribution? | Engagement scale score | Mean/SD or median/IQR; frequencies; CI where appropriate |
| Test association | Is support associated with engagement? | Support score; engagement score | Correlation and/or regression |
| Adjust for covariates | Does the association remain after adjustment? | Support, engagement, prespecified covariates | Multiple regression with diagnostics |