CY1500 Methodology by Zul Strait

Group 1

cy1500 — 2019-08-28 18:31:35 UTC

1. Control Group: People who eat outside home less than x times a week
Test Group: People who eat outside home more than x times a week

(Is this an expt or a "longitudinal study"?)

2. Big enough ~ 100 people

3. Hereditary factors, Nutritional values of the food eaten, physical activities, gender, ethnicity, BMI, age

4. Stratified random sampling

5. Measure blood glucose level, weight, weight and blood pressure level, participants to keep a food diary and record what they eat for every meal

6. Student's t-test

7. Participants should be allowed to receive treatment for diabetes (and drop out of the study?) if they develop the condition during the study

Group 2

cy1500 — 2019-08-28 18:31:40 UTC

1. Control: A group of people who never eat outside

Test group: Different groups of people who eat outside at different frequencies

2. Sample size: 3000

3. Type of food, Lifestyle, Fitness level, Genetics, Level of exercise, Age, gender, Eating portions, Race, What "else" they eat, Type 1/2, number, Different weight, Pre-existing medical conditions

4. Stratified Quota sampling

5. Over a duration of 1 year, make people report their eating details weekly. Test for risk for diabetes once a month (for example glucose level)

6. t-test

7. If our hypothesis is true, by controlling the number of times each participant eats out, we are putting the test group that eats at a high frequency at higher risk of diabetes

Group 3

cy1500 — 2019-08-28 18:31:43 UTC

1. control: people who only eat at home (+ve) people who only eat outside (-ve)
test: many test groups varying the number of meals they eat outside

2. 1000 people

3. number of meals per day, time of meal, type of food, activities that they do, BMI of people (within a specified range), quantity of food

4. age of test group, race of test group, gender, different basal metabolism

5. Systematic stratified random sampling

6. anova (p value < 0.05)

7. potential negative health impacts on subject groups, experiment is being done on actual human subjects instead of those cute white mice

Group 4

cy1500 — 2019-08-28 18:31:49 UTC

Q1.
Control - never eat outside home at all
test groups - people who eat 1 time, 5 time, 10 times a week etc

Q2.
Minimally 100 people to be credible.
Bigger sample size == more people to place into each test group == more concise result

Q3.
Age, gender, body composition, food nutritional value, lifestyle (eg. no. of exercise hours per week)

Q4.
When participants apply they will need to sign a form and declare their nationality, race

Q5.
Measure their waistline & BP, set them to eat for a set amount of time - then come back and re-measure their waistline & BP

Q6.
t-test for the statistical analysis

Q7.
Some people may get adverse health effects from eating restricted food everyday, which can be chronic.

Group 5

cy1500 — 2019-08-28 18:31:54 UTC

Q1. 4-5 test groups: 1 group eats out 1-5 times a week, 1 group eats out 6-10 times a week...etc., control group: don't eat out at all

Q2. Total 600 people, about 100 per group

Q3. Age, gender, amount of exercise, amount of sleep, amount of stress, preferably choose people who do not have a genetic disposition to diabetes.

Q4. Choose a sample with proportional race composition compared to the whole Singapore population. (e.g. 80% of sample should be chinese)

Q5. Get people in different test groups to eat outside/at home according to their test group for a year. Then, let them do the Diabetes Risk Assessment Quiz once at the end of every month for one year.

Q6. ANOVA Test

Q7. Forcing people to conform to a certain type of diet that they might not feel comfortable doing. The study itself might also lead to an increased risk of diabetes in the people eating outside many times a week.

Group 6

cy1500 — 2019-08-28 18:31:57 UTC

1. control group: people who only eats at home
test group: people who eats out 7 times a week, 14 times, 21 times (to simulate 1 meal, 2 meals and 3 meals per day)

2. sample size: 400 (100 per group)

3. family history of diabetes, age, gender, race, weight, risk factors, exercise frequency, socio-economic status, water

4. check the percentage of racial, age, gender, weight, etc. composition and match it accordingly to different groups; try to keep the proportion of participants as accurate as possible to the real SG population

5. ensure that they keep a nutrition and exercise journal over the span of the research (monitor once a week); conduct diabetes test every month for six months

6. ANOVA + Box and Whisker plot (to take into account mean, median mode and deviation)

7. loss of personal freedom in choice of food; deterioration of their health (for group that eats out); racial profiling