Collecting Data
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.
Data collection methods include surveys, observations, and experiments
The population is the entire group; a sample is a subset used for study
Random sampling helps ensure your sample represents the population
Good questions are clear, unbiased, and offer appropriate answer choices
Larger samples generally produce more reliable and accurate results
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.
Collecting Data
1 / 11
Collecting Data
Learn the different methods of collecting data and how to choose the right approach for your research question.