Designing a Survey

A school wants to know what new clubs students would join. How should they collect this data?

Define the population: All students at the school (let's say 500 students) = Population = 500 students

Choose a sampling method: Random sampling: Select 50 students using random number generator = Sample size = 50 (10% of population)

Design the survey question: Which of these clubs would you be interested in joining? (Check all that apply): Drama, Robotics, Art, Music, Debate = Multiple choice question with clear options

Plan data collection: Give survey during homeroom to avoid bias from time of day or class subject = Consistent collection method

Answer: Use a random sample of 50 students with a multiple-choice survey administered during homeroom.

Observation Study

A park ranger wants to count how many birds visit a feeder. Design an observation method.

Define what to measure: Count number of birds at the feeder at specific times = Variable = bird count

Set observation schedule: Observe for 15 minutes at 7 AM, 12 PM, and 5 PM each day for one week = 21 observation periods total

Create a recording system: Use a tally chart: Date | Time | Tally | Total = Organized data collection sheet

Establish consistency: Same person observes, same distance from feeder, count only birds on feeder (not nearby) = Standardized procedure

Answer: Observe at three fixed times daily for one week, using a tally chart to record counts consistently.

Designing an Experiment

A gardener wants to test if a new fertilizer helps plants grow taller. Design an experiment.

Define the variable: Measure plant height in centimeters after 4 weeks = Variable = plant height (cm)

Create treatment groups: Group A: 20 plants with fertilizer, Group B: 20 plants without fertilizer (control group) = 40 plants total, two equal groups

Control other factors: Same soil, same sunlight, same water amount, same plant type = Only fertilizer differs between groups

Plan data collection: Measure and record height of each plant at start and after 4 weeks = Before and after measurements for comparison

Answer: Use 40 identical plants split into treatment and control groups. Keep all conditions the same except fertilizer. Measure heights after 4 weeks.

Mistake: Using a convenience sample instead of a random sample

Why: Surveying only your friends or people near you creates bias. Your sample won't represent the whole population.

Correct: Use random selection methods: draw names from a hat, use a random number generator, or select every nth person from a list.

Mistake: Asking leading or biased questions

Why: Questions like 'Don't you agree that...' or 'How much do you love...' push people toward certain answers.

Correct: Use neutral wording: 'What is your opinion about...' or 'How would you rate...' with balanced options.

Mistake: Collecting too little data

Why: Small samples are unreliable. Flipping a coin 5 times might give 4 heads, but 100 flips will be closer to 50-50.

Correct: Aim for sample sizes of at least 30-50 for basic studies. Larger samples give more reliable results.

Medical Research

Pharmaceutical companies test new medications using carefully designed experiments called clinical trials.

To test a new headache medicine, researchers randomly assign 1000 patients to receive either the new medicine or a placebo. They collect data on pain relief, side effects, and recovery time.

Market Research

Companies use surveys and focus groups to understand customer preferences before launching products.

A phone company surveys 2000 randomly selected customers to find out which features matter most: camera quality, battery life, screen size, or price.

Q: What is the difference between a census and a sample?

A: A census collects data from every member of the population. A sample collects data from only part of the population. Censuses are more accurate but expensive and time-consuming. Samples are practical when populations are large.

Q: How big should my sample be?

A: There's no single answer, but generally: at least 30 for basic analysis, 100+ for better reliability, and 1000+ for high precision. The needed size depends on how much variation exists in your population and how precise you need to be.

Q: What makes a question biased?

A: A question is biased if it encourages a particular answer. Leading words (obviously, surely), loaded terms (wasteful, amazing), or unbalanced options all create bias. Always test your questions with others before using them.