1 of 19

Teaching Introductory Statistics with ISI

bchance@calpoly.edu

Beth Chance

Cal Poly – San Luis Obipso

2 of 19

Example from my course

3 of 19

Stat 217 – Day 18

Comparing Two Proportions (Ch. 5)

4 of 19

Last Time – Comparing Groups (Ch. 4)

  • Identifying explanatory variable and response variable
    • Example: Many studies have shown that women who smoke while pregnant tend to have babies who weigh significantly less at birth, on average, than women who do not smoke while pregnant
      • Explanatory: whether or not mother smokers (categorical)
      • Response: baby’s birth weight (quantitative)
    • Remember to focus on what you measure about each observational unit
      • Do babies weigh significantly less – categorical x
      • Baby weight – quantitative
    • Research question: Are the two variables associated?

5 of 19

Last Time – Comparing Groups (Ch. 4)

  • Type of study
    • Observational study – observed what happens naturally
    • Experimental study – active imposition of EV
    • Example: Many studies have shown that women who smoke while pregnant tend to have babies who weigh significantly less at birth, on average, than women who do not smoke while pregnant. Could this have been an experiment?
      • Only if they randomly decided which mothers would smoke

​

6 of 19

Last Time – Comparing Groups (Ch. 4)

  • Confounding variables
    • A third variable that changes with the explanatory variable and could provide an alternative explanation for the observed association
    • What else might differ between children with smoking moms and non-smoking mom’s that might also be related to birth weight?
    • (as a group tendency, not just one mom)

​

​

7 of 19

Last Time – Comparing Groups (Ch. 4)

  • Randomized comparative experiment
    • Goal of Random Assignment – create groups that are similar to each other on all other variables
    • If the groups were “equivalent” before we imposed our explanatory variable and then we see a difference, can draw a cause-and-effect conclusion

​

8 of 19

Ch. 4 Review

  • With a randomized, comparative experiment have the potential to draw a cause-and-effect conclusion between the explanatory and response variable.
    • Experiment = active imposition of explanatory variable
    • Comparative = at least 2 groups
    • Randomized = random assignment
    • Double-blind = neither experimenter or evaluator of response variable knows which group subject is in

(placebo effect)

9 of 19

Best Figure in Entire Course

10 of 19

Uses of randomness

11 of 19

Comparing two proportions

  • Two-way table
  • Graphical summary: Segmented bar graph or mosaic plot
  • Numerical summary:
    • difference in conditional (EV) proportions
  • Looking for evidence of an association between response variable and explanatory variable
    • But how do I know I didn’t just get an unlucky random assignment, just by chance alone?

​

12 of 19

Lab 4

  • Prelab
    • Could random assignment have plausibly created the difference between the groups?
    • Need to know what random assignment looks like
      • How larger of a difference between the two groups might I expect to see from random chance alone?
      • Is the difference between the two groups larger than what random assignment might created with all the success/failure outcomes “fixed”

​

​

​

  • Identify the observational units
  • Identify the explanatory and response variables
  • Experiment or observational study
  • Potential to draw cause-and-effect conclusion?

13 of 19

Lab 4

14 of 19

Powerpoints

15 of 19

To Do

  • Lab 4 (due Tuesday)
  • Review for Quiz Monday night (Ch. 4)
  • Optional reading in Canvas: Ethics

16 of 19

Assessment

After Wed class

After Thur class

​

​

  • Before Monday

17 of 19

WileyPlus

18 of 19

WileyPlus

19 of 19

WileyPlus

CourseWide Resources