Investigating what health means to young people: question, method, ethics and findings

Most weak investigations are weak before any data is collected, because the question was never answerable. Four tests catch nearly all of it.

• Is it answerable with data you can actually collect? “Are young people healthy?” is not; “What do Year 11 students at this school say health means to them?” is.
• Is it specific about who and what? Name the group and the thing you are asking about. “Young people” is not a group you can sample.
• Is it neutral? A leading question contains its own answer. “Why do young people ignore their mental health?” has already decided that they do.
• Would different answers be interesting? If every plausible answer leads to the same conclusion, the question is decorative.

Notice what the syllabus dot point asks: the meanings of health for young people. That is a question about meaning, so the answers will be in words, and a question about meaning cannot be answered honestly by a five-point scale alone.

The syllabus names three: survey, interview questions and focus groups. Each answers a different shape of question, and choosing on convenience rather than on fit is the commonest methodology mark lost.

For a question about meaning, a survey alone is the weakest choice, because you must write the options in advance and the options then become the answer. A common strong design is a mixed method: a short open-response survey to find the range of meanings, then interviews to understand a few of them properly.

Two sampling points worth a mark each. Your sample is not your population, so say who you actually reached. And a sample collected from people who volunteer is not the same as one collected at random — the students most interested in health are the likeliest to volunteer, which biases the answer in a direction you can name.

Research about health asks people about their bodies, their families and sometimes their worst weeks. The ethical requirements exist because that can do harm, and they apply to a Year 11 investigation exactly as they apply to a university one.

• Informed consent. Participants know what the study is, what will happen to their answers, and that they may stop at any time without giving a reason. For participants under 18, consent normally involves a parent or guardian as well.
• Voluntary participation. Nobody is pressured. A researcher who is also a friend, a captain or a teacher carries influence, and that influence has to be actively neutralised rather than assumed away.
• Anonymity or confidentiality. Decide which you are offering, say so plainly, and do not promise anonymity if you can identify respondents by their handwriting or their class.
• No harm. Questions about mental health, body image, food or family can cause distress. Have a plan for that, including where to direct someone who needs support, and never present it as your job to provide the support yourself.
• Cultural safety. Research involving Aboriginal and Torres Strait Islander people carries additional obligations of community consultation and control. Do not treat a cultural group as a data source.
• Honest reporting. Report what you found, including the answers that spoiled your expected pattern. Quietly dropping them is falsification.

These three get used as though they were synonyms for “good”. They are not, and an answer that keeps them separate is immediately visible as a stronger one.

• Validity asks: did you measure what you meant to measure? If your question is about meanings of health and your survey counts how many times a week people exercise, the data may be perfectly accurate and completely invalid for your question.
• Reliability asks: would you get the same result again? If two interviewers word the questions differently, or you ask a group right after a wellbeing assembly, the result moves for reasons that have nothing to do with the thing you are studying.
• Credibility asks: can this source be trusted? Who collected it, what did they stand to gain, is the method described well enough to check, and does it agree with other sources.

They are independent, which is the point. A study can be reliable and invalid — consistently measuring the wrong thing. It can be valid but unreliable — asking exactly the right question in such an inconsistent way that the answer wobbles. And a study can be both and still lack credibility if the people who ran it had an interest in the result.

Two limits that apply to a school investigation and should be stated rather than hidden: a small sample cannot support a claim about all young people, and a sample drawn from one school describes that school. Saying so is a strength. Claiming more than your data carries is the single commonest way to lose marks in a conclusion.

A finding is what your data shows. A conclusion is what you are entitled to say because of it. Keeping them apart is most of the skill.

Present qualitative findings by grouping responses into themes and saying how many responses sat in each, quoting one real response per theme. “Eleven of twenty-four described health mainly as being able to keep up with what their day demands” is a finding. “Young people define health functionally” is a conclusion, and a bigger one than twenty-four students can support.

Now bring Lesson 5 back, because the syllabus asks it explicitly: the determinants of health shape what a young person counts as health. If your respondents from one part of town answer differently from another, that is not noise. Income, housing, work, culture and access to services change what health has cost a person and therefore what they notice about it. Reading a difference as patterned rather than random is what turns a list of quotes into analysis.

Finally, identify the next research question. This is a dot point in its own right and takes one sentence: name something your findings raised but could not answer. A study that answers everything it asked and suggests nothing further has usually asked something too small.

Each design below has one dominant flaw. Name whether it damages validity, reliability, credibility or ethical conduct, justify it in a sentence, and propose the specific repair.

Build a complete one-page proposal for investigating what health means to young people in your community. It must contain all seven of these, each in no more than two sentences:

• The research question, in its final form.
• The method, with a justification beginning “I chose this because the question asks for …”.
• Who you would collect from, and how, and how that differs from the population you want to describe.
• Three questions you would actually ask, at least two of them open.
• How you handle consent, withdrawal and distress.
• One threat to validity and one to reliability that your design cannot remove.
• The further research question you expect it to raise.

Then run it once, on one person, using your three questions exactly as written. Record their answers in their own words. Afterwards, note the one thing you would change about your questions — a question that worked on paper and not in the room is the most useful finding a pilot produces.