MYP Design Presenting Research and Testing Evidence Notes

Design the Page for Someone Skimming It

1. A reader gives a research page a few seconds before deciding to read it properly, so put the finding where the eye lands first.
2. Head every page with the finding itself, like "Users wash the bottle less than once a week", instead of the word "Questionnaire".
3. Show the evidence underneath the heading, then close with one line naming what you changed because of it.
4. Stick to one or two columns, left aligned text and the same body font size on every page of the folder.
5. Empty space around a chart is doing work, because a cramped page hides the finding you wanted read.
6. Caption every chart and photograph, since captions get read when body text does not.

Summarise a Questionnaire, Never Paste It

1. Thirty completed questionnaires are raw data, and raw data on its own is not a finding.
2. Count the answers, turn the counts into one chart or a small table, then write the two or three findings that matter.
3. State the sample plainly, for example "24 of 30 year 8 students replied".
4. Give every number its total, because "18 out of 24" tells a reader far more than "most".
5. Quote one or two real sentences from the open questions, since a user's own words carry something a bar chart cannot.
6. Finish with the design consequence, such as dropping the screw lid after 19 of 24 said they open the bottle one handed.

Tables and Charts a Reader Can Follow

1. Bar charts compare categories, line graphs show change over time, and a pie chart works only for parts of one whole with a few slices.
2. Label both axes with what is measured and the unit.
3. Put the sample size beside the chart, because a bar chart of 5 responses looks identical to one of 500.
4. Give one chart one message, and split any chart that needs two sentences of explanation.
5. Leave out 3D effects, drop shadows and rainbow colour schemes, which distort the bars and make close values harder to compare.
6. Use a table when exact values matter, for example five materials with cost per sheet, thickness in millimetres and mass.

Annotation Is What Turns a Photograph Into Evidence

1. A photograph on its own records what something looked like, and the annotation is what makes it say something.
2. Add arrows or numbered labels pointing at the exact feature you are discussing, the hinge, the glued joint, the burn mark.
3. Write one line saying what the shot proves, for example "the corner lifted about 3 mm off the print bed".
4. Include something of known size in the frame, a 30 cm ruler or a coin, so a reader can judge scale.
5. Shoot against a plain background in even light, because a cluttered bench hides the detail you came to photograph.
6. Pair a before and an after taken from the same angle and distance, since two shots from different angles prove nothing.

Test Tables That Prove Something

1. Every test should trace back to a numbered specification point, so make that the first column of the table.
2. Record what you did, what you expected, what you measured and what you will change.
3. Write real units and real numbers, so "held 4 minutes 20 seconds before the tape peeled" beats "held well".
4. Repeat each test at least twice and record both results, because a single reading can be luck.
5. Leave the failed tests in, since a failure with a stated fix is worth more than a column of passes.

Raw Data Lives at the Back

1. The appendix holds the full questionnaire responses, the interview transcript, every repeat measurement and any long code listing.
2. Number the appendices and call them by number from the main page, like "full responses in Appendix 2".
3. Nothing that earns marks should live only in an appendix, because a reader may never view it.
4. The appendix is what makes your summary believable, since anyone doubting a chart can go and check the numbers on it.
5. Keep appendix pages plain and spend no styling time on them, because their job is completeness.