The Mindfulness Audio and the Two Intact Sections
Sharada, counsellor at a Karnataka pre-university college, wants to know whether a 10-minute mindfulness audio played before first period improves attention in I PUC. She has two intact Commerce sections of 48 students each, and the master timetable will not permit moving students between them. Section A's first period is Psychology; Section B's is Accountancy. Section A already has better attendance and a higher internal average. Her plan: give both sections a 20-item attention test in June, play the audio for Section A daily for six weeks, and retest both sections in August. The principal wants a finding before the October internal exams and asks whether Sharada can simply 'announce it to parents as an experiment' and tell students the audio is 'just relaxing music' so that they do not try harder.
1. Sharada claims she is running an experiment. Evaluate that claim and state, with justification, what design she can actually use within her constraints.
Read the worked answer
She is not running a true experiment. The defining feature of an experiment is random assignment, which creates groups presumed equivalent before treatment. Sharada's groups are intact sections, so any difference may reflect pre-existing differences (attendance, prior achievement) or the different subject taught in first period. What she has is a quasi-experimental, non-equivalent control group design with pre- and post-test.
How to make the best of it:
- Report it as quasi-experimental, never as experimental; avoid causal verbs.
- Use the June pre-test as a covariate (or match pairs of students on pre-test scores), so post-test differences are adjusted for starting level.
- Remove the subject confound by measuring attention immediately before and after the audio within the same first period, or by running an ABAB sequence in Section A (audio weeks alternated with no-audio weeks), so the section partly serves as its own control.
- Keep tester, timing, instructions and room conditions constant across both sections.
Conclusion: she can produce useful, cautious evidence ('improvement was associated with the audio, after adjusting for pre-test scores'), but not a causal claim, because selection was not under her control.
2. Name two threats to internal validity in her plan and one concrete safeguard for each.
Read the worked answer
- Selection - Section A differs before treatment (better attendance, higher internal average), so a post-test difference may simply reproduce that gap. Safeguard: pre-test covariate or matched pairs, plus the ABAB reversal; compare change scores rather than raw post-test means.
- Testing / practice effect - the same 20-item test in June and August lets students improve from familiarity with the items, not from attention. Safeguard: parallel forms of equal difficulty, with form order counterbalanced across sections.
- Expectancy and novelty (Hawthorne effect) - students know something new is happening and may perform differently; Sharada also scores the tests knowing who received the audio. Safeguard: blind scoring by another teacher, a standardised script, and a 'placebo' audio (quiet instrumental) for Section B so both groups expect something.
- Maturation and history - eight weeks includes ordinary development and August exam stress. Safeguard: shorten the interval and record events that could affect both groups.
Any two explained threats with a workable safeguard would earn full marks; merely naming them does not.
3. Assess the ethics of the two disclosures Sharada proposes: calling the audio 'just relaxing music', and 'announcing it to parents as an experiment'.
Read the worked answer
Deception. Describing the audio as 'just relaxing music' withholds the purpose. Deception is defensible only when it is necessary, risk is minimal, the right to withdraw is intact, and participants are debriefed afterwards, told the real aim, and allowed to withdraw their data. Here the deception also damages the study: students who suspect a test behave differently, so the ethical and methodological objections point the same way.
Consent with minors. I PUC students are usually 16, i.e. minors. The principal's permission is institutional approval, not guardian consent; written information to parents with an opt-out, plus the students' own assent, is required. A single announcement is not consent.
Fairness and confidentiality. If the audio looks beneficial, Section B should receive it after the study (wait-list design) so no one is denied a possibly useful practice. Individual attention scores must not be shown to subject teachers in identifiable form, participation must carry no academic penalty, and students must be able to stop attending without explanation. 'Announcing it as an experiment' to parents satisfies none of these by itself.