GUESS THE METHOD
Researchers DMed 3,000 TikTok users asking how many hours per day they doom-scroll and their self-reported anxiety level. Nobody's FYP was touched. No variables manipulated.
Correlational study — two variables (doom-scrolling time and anxiety) were measured in the same participants to find a relationship. Nothing was manipulated.
Five friends rated how 'delulu' they were on a scale of 1–10 after a rough Monday. What is the mean? What is the mode?
Ratings: 6, 9, 9, 4, 7
ANSWER:
Mean = (6+9+9+4+7) ÷ 5 = 35 ÷ 5 = 7. Mode = 9 (appears twice — clearly the most delulu friend group).
A scatterplot shows hours of daily BeReal + Instagram posting (x-axis) vs. self-reported loneliness score (y-axis). Dots trend upward left to right, loosely clustered (r ≈ +0.55). Describe the direction, strength, and whether we can blame social media.
ANSWER:
Moderate positive correlation: more posting is associated with higher loneliness. We CANNOT say social media causes loneliness — lonely people may post more to cope. Correlation does not equal causation.
Analyze this data on hours of sleep vs. how often students said 'I'm literally so tired' in a day. What pattern exists? Name one limitation.
4 hrs sleep: 11.2 complaints/day 5 hrs sleep: 8.7 complaints/day 6 hrs sleep: 5.4 complaints/day 7 hrs sleep: 3.1 complaints/day 8 hrs sleep: 1.8 complaints/day
Strong negative pattern — as sleep increases, complaints decrease dramatically (11.2 down to 1.8). Limitation: self-reported data. Students may overclaim tiredness for social clout, making this unreliable. Also correlational — can't prove sleep causes fewer complaints.
A researcher wants to test if getting a 'W' reaction on your BeReal improves self-esteem. She posts a BeReal for 50 students and checks their self-esteem the next day. What critical design flaw exists?
No control group and no random assignment. Some students got W reactions naturally and some didn't — but this wasn't controlled. She also can't rule out other events that day affecting self-esteem. Solution: randomly assign students to receive a W reaction vs. no reaction condition.
Correlational study — two variables (doom-scrolling time and anxiety) were measured in the same participants to find a relationship. Nothing was manipulated.
Naturalistic observation — researchers recorded real behavior in its natural setting (the cafeteria) without interfering or manipulating anything.
Six students tracked how many times they checked their phone during a 50-minute class. Find the median and the range.
Phone check counts: 34, 12, 67, 19, 22, 41
ANSWER:
Sorted: 12, 19, 22, 34, 41, 67. Median = (22+34) ÷ 2 = 28 checks. Range = 67−12 = 55 checks. Someone is not okay
Two scatterplots show hours of sleep (x) vs. AP exam score (y). Plot A has tight clustering near the line. Plot B has scattered dots with a slight upward trend. Which shows a stronger correlation? What does the positive direction mean?
ANSWER:
Plot A is stronger (r ≈ +0.97, close to 1.0). Plot B (r ≈ +0.45) is weaker. Positive direction = more sleep is associated with higher AP scores in both cases.
Does the data support the hypothesis that the Duolingo owl's guilt-tripping actually works?
Duolingo with guilt-trip owl: 47 words learned/week Duolingo without owl: 31 words/week Quizlet only: 28 words/week Did nothing: 9 words/week
Yes — the guilt-tripping owl condition (47 words) outperformed every other condition by a wide margin. The owl's passive-aggressive energy appears effective. However, random assignment and statistical significance testing are needed before claiming causation.
Identify the IV, DV, and one confounding variable in this study.
90 high schoolers were randomly assigned to watch a 10-min Sabrina Carpenter music video, a 10-min lo-fi playlist, or sit in silence before a vocab quiz. Quiz scores were recorded.
IV: Pre-quiz condition (Sabrina / lo-fi / silence). DV: Vocab quiz score. Confound: Pre-existing preference for or emotional connection to Sabrina Carpenter — fans might feel more energized and motivated regardless of which condition they were in.
Naturalistic observation — researchers recorded real behavior in its natural setting (the cafeteria) without interfering or manipulating anything.
Meta-analysis — combines and statistically re-analyzes data from many existing studies to reach broader conclusions. Pooling 58 studies gives far more statistical power than any single study.
The average 'main character energy' score for 25 students was 6.4 out of 10. Then one student who scored a perfect 10 joined the class. How did the mean change?
Original: n=25, mean=6.4 | New student score: 10
ANSWER:
Original sum = 25×6.4 = 160. New sum = 160+10 = 170. New mean = 170÷26 ≈ 6.54. The one main character raised the whole class average by 0.14 points.
A bar graph shows average 'rizz score' (self-reported confidence, 1–10) across five social situations. Analyze the pattern — does rizz peak in the DMs? Does the graph prove situation CAUSES confidence?
ANSWER:
Rizz peaks in the DMs (7.8) and drops sharply for IRL situations — especially class presentations (3.1). Social context strongly affects confidence. This is observational data — it does NOT prove the situation causes the confidence level.
Evaluate this hypothesis: 'Students who listen to music while studying will score higher on tests than those who study in silence.'
Music (any genre): mean 74.2, SD 8.1 Silence: mean 76.8, SD 7.4 Lofi hip hop only: mean 78.3, SD 6.9 Chaotic playlist: mean 68.5, SD 12.2
NOT fully supported. General music (74.2) actually scores LOWER than silence (76.8). Lofi hip hop (78.3) is the exception — it beats silence. A chaotic playlist (68.5) actively hurts performance. The TYPE of music matters enormously. High SD in the chaotic group means wildly inconsistent results.
A study on whether 'rizz' can actually be taught reports p = 0.03. What does this mean? What does it NOT mean?
p = 0.03 means only a 3% chance the results occurred by random chance — we reject the null hypothesis. It does NOT mean rizz is definitely teachable, the effect size is large, or the finding is practically meaningful. A huge sample can make a trivially small difference statistically significant.
Dr. Chen spent 2 years intensively documenting every detail of a single teenager who went viral overnight — tracking his mental health, sleep, relationships, and behavior as his follower count hit 10 million.
Case study — an in-depth investigation of ONE individual over time. Very detailed but not generalizable to all viral teens. Great for rare or unique situations.
Which group of Spotify listening habits has a higher standard deviation, and why?
Group A (hyperpop fans): 180, 182, 179, 181, 183 min/day Group B (the aux cord enjoyers): 40, 95, 180, 220, 310 min/day
ANSWER:
Group B has a MUCH higher standard deviation. Group A's listening times are nearly identical (SD ≈ 1.5 min). Group B has values scattered wildly from 40 to 310 — huge deviations from the mean of 169 min.
A line graph tracks GPA over 4 semesters for three groups: 0 hrs/week gaming, 5–10 hrs/week, and 15+ hrs/week. What story does the data tell? What limits our conclusions?
Heavy gamers (15+ hrs) start lower and decline sharply over time. Moderate gamers hold steady. Non-gamers stay highest and stable. This correlational data suggests heavy gaming is associated with GPA drops — but we can't BLAME gaming without a controlled experiment.
A researcher hypothesized that texting speed correlates with GPA. Analyze the results. Is the hypothesis supported? What would strengthen the study?
<20 texts/min: avg GPA 3.4 (n=45) 20–40 tpm: avg GPA 3.2 (n=88) 40–60 tpm: avg GPA 2.9 (n=76) 60+ tpm: avg GPA 2.6 (n=41)
ANSWER:
There IS a negative pattern — faster texting correlates with lower GPA. But this is correlational and likely reflects overall phone usage habits. To strengthen: randomly assign phone use levels, control for other variables, and run statistical significance tests to get a p-value.
Researchers study whether people who use 'NPC' as an insult score lower on empathy tests. They post a survey on Reddit. Identify TWO threats to validity.
(1) Sampling bias — Reddit users skew younger, more online, and less demographically diverse than the general population. Results can't generalize broadly. (2) Social desirability bias — people may not honestly admit to using mean insults or report low empathy on a self-report survey.
To test whether Celsius energy drinks actually improve reaction time, 200 students were randomly assigned to drink Celsius, a placebo drink, or water before a gaming reaction test. Scores were compared.
Experiment — the researcher manipulated the IV (drink type), used random assignment to conditions, and measured the DV (reaction time score). The ONLY method that can establish causation
Charli XCX's 'Brat' album dropped and anxiety scores in a survey of 400 teens were recorded. One student scored at the 92nd percentile. What does that mean? Does it tell us their actual score?
ANSWER:
It means this student scored higher than 92% of all 400 teens surveyed. Percentile rank = relative standing in a distribution. It tells us NOTHING about the raw anxiety score — just how they compare to the group.
A scatterplot shows skip-school TikToks watched per week (x) vs. actual absences per month (y). There is a strong positive correlation. A classmate says 'TikTok MAKES kids skip school.' What is wrong with this? Name a confounding variable.
Correlation does not imply causation. Students who are already disengaged from school may both watch these TikToks more AND skip more — pre-existing school disengagement is the confounding variable. The TikToks don't necessarily cause the absences.
What pattern do you see in this data on Stanley cup refills vs. productivity? Name TWO confounding variables that could explain it.
0 refills: avg productivity 4.1/10 1 refill: avg productivity 5.8/10 2 refills: avg productivity 7.2/10 3+ refills: avg productivity 7.6/10
Positive pattern — more hydration correlates with higher productivity (4.1 to 7.6, leveling off). Confound 1: Health-conscious students may both hydrate more AND be more productive overall. Confound 2: Students at well-resourced schools may have easier water access AND better academic support simultaneously.
A journalist reports: 'Students who drink more Celsius get BETTER grades — energy drinks boost academic performance!' They cite r = +0.41 between Celsius cans/week and GPA. List THREE problems with this headline.
(1) Correlation does not equal causation — r = +0.41 cannot support a claim that Celsius BOOSTS grades. (2) Confounding variable — wealthier students may afford both more Celsius AND better tutoring. (3) r = +0.41 is a moderate correlation — it explains only about 17% of GPA variance. The headline wildly overstates the finding.