S2 Ep2 Puzzling Over Populations and Samples

Data’s Anatomy: Season 2, Episode 2, “Puzzling Over Populations and Samples”

Narrator: Office of Institutional Intelligence and Research (IIR) Presents:
Data’s Anatomy
Narrator: Puzzling Over Populations and Samples Season 2: Episode 2
Narrator: Featuring….
Narrator: Dan & Daphne: Data Interns

Narrator: On this episode…
Narrator: Topic: Populations vs. samples
Narrator: Importance: Understanding the difference between what constitutes a population versus a sample
Narrator: When to use or not use statistical significance with these groups

Narrator: One morning in IIR.

Dan: Hey Daphne, I just got assigned a TCCD FTIC comparison report.
Daphne: What comparison?
Dan: 2014 FTIC to 2015 FTIC…
Dan: I’ll show you when all the statistics are done…
Daphne: All the statistics?
Daphne: This oughta be rich!

Narrator: The next day…

Daphne: Good morning, Dan!
Daphne: Gosh, you look terrible!
Dan: I stayed up here all night running T-test statistics on the FTIC success outcomes.
Daphne: Oh no!
Dan: Worst of all, I didn’t even get statistical significance!
Dan: Maybe I need a bigger sample…
Daphne: Well, cheer up buttercup! I have some good news…

Daphne: Dan, if you’re working with the FTIC groups, you don’t have a sample…
Daphne: That’s the entire FTIC population!

Daphne: Social scientists use statistical significance in order to make an argument…
Daphne: Although researchers want to study a certain population, they often have limitations that do not allow them to observe everyone.
Daphne: In that case, the researcher can take a sample of that population…
Daphne: The researcher would then use statistical significance to make inferences about the population based on the data collected from the sample.

Daphne: So, if you have an entire population, you don’t need statistical significance.
Daphne: Therefore you can just look for differences in descriptive statistics such as mean, median, and distribution.
Dan: That’s too bad, though. I really wanted to get statistical significance to support my point!
Daphne: You got it all wrong, Dan!
Daphne: Having the entire population is as good as it gets!

Daphne: For social science researchers, it is usually impossible to study an entire population.
Daphne: For example, maybe you population of interest is all Texans.
Daphne: This population would be defined as those who live within the state boundaries and are residents.
Daphne: Trying to study this whole population would be time-consuming, costly, and unrealistic.
Daphne: So you would have to take a sample instead…
Daphne: In this case, researchers would need statistical significance to make inferences about the population based on the data collected from the sample.
Daphne: However, if you can feasibly measure an entire population, it’s best.

Daphne: At TCCD, our population is defined as students who are registered/enrolled.
Daphne: This means we can almost always measure our entire population.

Daphne: And this means that we don’t need statistical significance and that we can make strong arguments based on population data!

Dan: Wait a minute!
Dan: That means that I didn’t have to spend all those hours running statistics on the TCCD FTIC report!!
Dan: Crikey! I’m gonna need therapy or something…
Daphne: Suddenly I’m really glad I didn’t go into counseling!

Narrator: To be continued…
Narrator: Next time of Data’s Anatomy: Studying Statistical Significance
Dan and Daphne: This is Dan and Daphne signing out…
Dan: Gotta run! See you next time!
Narrator: IIR – Slamming the Data!
Dan: For more information go to our website.
Narrator:
Dan: Be sure to check our existing reports and other helpful videos.
Narrator: Questions: Contact us at 817-515-1516

Samples