The training is over, the surveys are collected and the report is ready. Yet an important question remains: what should change at work? This happens in education, staff development and programme evaluation. Findings become more useful when they address a clear question and inform an action the team can carry out.
Begin with the decision to support
Before selecting a survey or test, establish what you need to know and why. Does the team need a new skill? Are instructions unclear? Are you considering a programme change or identifying people who need support? A clear question guides the indicator, instrument and evidence source, keeping data collection relevant to the task.
For example, determining whether a staff member can perform a task may require a work sample or observation alongside their feedback on training. Satisfaction with a session is useful information, but judging application requires evidence connected to the task itself.
Examine the data before interpreting it
Review who participated and who did not, the sampling approach, question clarity, missing values and coding. Ask whether the instrument gives participants a fair opportunity to demonstrate what they know, and whether language or administration conditions influenced the findings.
Statistical methods depend on the research question, design, variable types, sample size and analytical assumptions. Start with an understandable description of the data, then consider whether further inference is appropriate. Documenting these choices helps readers understand both the findings and the limits of what can be concluded.
Separate observation from interpretation
“Participants struggled with the task” is an observation whose supporting evidence needs to be specified. “The training was unsuitable” is an interpretation to examine. The difficulty might relate to content, time for application, staff authority or the clarity of a procedure.
A difference between two measurements deserves similar scrutiny. Did the same people participate? Were the instruments and conditions comparable? Could other factors explain the difference? These questions help keep conclusions proportionate to the design; an association alone does not establish causation.
An illustrative example: a team missing task deadlines
Suppose an organisation wants to address delays in a recurring task. A training course might be proposed immediately. First, it could review the task steps, listen to the team and examine suitable work samples with personal data removed.
If the evidence points to overlapping responsibilities, clarifying roles and approvals may be the priority. If it reveals a skill gap, focused practical training may help. Both actions might be needed. This is an illustration of the reasoning process, not a completed project result or a diagnosis that applies to every team.
Compare options and choose a manageable action
Identify options worth discussing, such as revising a procedure, providing training or testing a support tool. Compare their relevance to the problem, resource needs, timing, constraints and unanswered questions. Where a decision forms part of wider planning, evidence can inform SWOT analysis and TOWS can help develop strategic options.
Once an option is chosen, record the finding, interpretation, action, responsible person, progress indicator and review date. A limited trial can be adjusted in response to evidence before wider adoption.
What this means at TAS
This approach connects assessment-instrument review, study and programme data analysis, training-needs assessment, team development and institutional decision support. The scope follows a review of the question and available inputs, including the limits of the findings and any need for additional information.
Start with one question: which decision is this report supposed to help me make? Then try the evaluation-design map, assessment-quality checklist or results-to-decisions template. Analysis earns its value when it helps us choose an understandable action and review what happened afterwards.