1. Introduction to Statistics for the Social Sciences SBS200 - Lecture Section 001, Fall 2017 Room 150 Harvill Building 10: :50 Mondays, Wednesdays & Fridays.
Welcome

2. writing assignment forms notebook and clickers to each lecture
Remember bring your
writing assignment forms
notebook and clickers to each lecture

3. Three tries
Will show up in grades within 24 hours

4. Lab sessions Labs start Today Everyone will want to be enrolled
in one of the lab sessions
Labs start Today

5. Project 1

6. Project 1 Likert Scale (summated scale) Correlation (scatterplots) Comparing two means (bar graph)

7. Likert Scale is always a “summated scale” with multiple items.
All items are measuring the same construct. The score reflects the sum of responses on a series of items.
- miniquiz (like Cosmo - ask several questions then sum responses)
- For example, several questions on political views (coded so that larger numbers mean “more liberal”)
1. Lower taxes and a smaller
government will improve the standard of living for all.
agree disagree
2. Marriage should be between one man and one woman
agree disagree
3. Evolution of species has no place in public education
agree disagree

8. I prefer rap music to classical music Agree 1---2---3---4---5 Disagree
Likert Scale is always a “summated scale” with multiple items.
All items are measuring the same construct. The score reflects the sum of responses on a series of items.
Anchored rating scales: a written description somewhere on the scale
I prefer rap music to classical music
Agree Disagree
Fully anchored rating scales: a written description for each point on the scale
I prefer rap music to classical music
Strongly Disagree
Strongly Agree
Agree
Neutral
Disagree

9. Likert Scale is always a “summated scale” with multiple items.
All items are measuring the same construct. The score reflects the sum of responses on a series of items.

10. Project 1 Likert Scale (summated scale) Correlation (scatterplots) Comparing two means (bar graph)

11. Scatterplot displays relationships between two continuous variables
Correlation: Measure of how two variables co-occur and
also can be used for prediction
Range between -1 and +1
The closer to zero the weaker the relationship
and the worse the prediction
Positive or negative

12. Positive correlation: as values on one variable go up, so do values
Positive correlation: as values on one variable go up, so do values for the other variable
Negative correlation: as values on one variable go up, the values for the other variable go down
Height of Mothers
by Height of Daughters
Height of Mothers
Positive
Correlation
Height of
Daughters

13. Positive correlation: as values on one variable go up, so do values
Positive correlation: as values on one variable go up, so do values for the other variable
Negative correlation: as values on one variable go up, the values for the other variable go down
Brushing teeth
by number cavities
Brushing Teeth
Negative
Correlation
Number Cavities

14. Perfect correlation = +1.00 or -1.00
One variable perfectly predicts the other
Height in inches
and height in feet
Speed (mph)
and time to finish race
Positive
correlation
Negative
correlation

15. Correlation
Range between -1 and +1
+1.00 perfect relationship = perfect predictor
+0.80 strong relationship = good predictor
+0.20 weak relationship = poor predictor
0 no relationship = very poor predictor
-0.20 weak relationship = poor predictor
-0.80 strong relationship = good predictor
-1.00 perfect relationship = perfect predictor

16. Correlation

17. Correlation - How do numerical values change?
Correlation - How do numerical values change?
Let’s estimate the correlation coefficient for each of the following
r = +.80
r = +1.0
r = -1.0
r = -.50
r = 0.0

18. Project 1 Likert Scale (summated scale) Correlation (scatterplots) Comparing two means (bar graph)

22. Final results might look like this
Predicting One
positive correlation 15 12 9 6 3
“Passion for Gaming” Score
Time Studying

25. Final results might look like this
Predicting One
positive correlation 15 12 9 6 3
“Passion for Gaming” Score
Time Studying

27. Final results might look like this
Predicting One
negative correlation 15 12 9 6 3
“Passion for Gaming” Score
Age

30. Final results might look like this
Predicting One
negative correlation 15 12 9 6 3
“Passion for Gaming” Score
Age

32. Final results might look like this
Predicting One
Group has bigger mean 15 12 9 6 3
“Passion for Gaming” Score
Gender
Female Male

33. Final results might look like this
Predicting One
Group has bigger mean 15 12 9 6 3
“Passion for Gaming” Score
Gender
Female Male

34. Final results might look like this
Average of three scores for males
Final results might look like this 10
Predicting One
Group has bigger mean 15 12 9 6 3
“Passion for Gaming” Score
Gender
Female Male

35. Final results might look like this
Average of three scores for females
Final results might look like this 12
Predicting One
Group has bigger mean 15 12 9 6 3
“Passion for Gaming” Score
Gender
Female Male

36. Final results might look like this
Predicting One
Group has bigger mean 15 12 9 6 3
“Passion for Gaming” Score
Gender
Female Male

37. Final results might look like this
Predicting One
Group has bigger mean 15 12 9 6 3
“Passion for Gaming” Score
Gender
Female Male

39. Final results might look like this
Predicting One
Group has bigger mean 15 12 9 6 3
“Passion for Gaming” Score
Gender
Female Male

40. “Serious Gamer” Score “Serious Gamer” Score “Serious Gamer” Score Time
One positive
correlation
One negative
correlation
Comparing
Two means
“Serious Gamer” Score
“Serious Gamer” Score
“Serious Gamer” Score
Time
Studying
Age
Gender

45. Project 1 - Likert Scale - Correlations - Comparing two means (bar graph)
Questions?

46. Random assignment Independent and Dependent Variables
How do we decide who gets into each condition?
Random assignment of subjects into groups:
Any subject had an equal chance of getting
assigned to either condition
Gender and spending
If random assignment
then you may have a “true” experiment
If no random assignment
then you have a
“quasi” experiment
Sleep
and memory
Music
and plants
Cell phones
and driving
Cars
and cool

47. What if we can’t randomly assign people to groups?!??
Random assignment of subjects into groups: Any subject had an equal chance of getting assigned to either condition
Comparing heights of 7-year-olds with 17-year-olds
Quasi-experiment: comparing “subject variables”
- no random assignment -
Height
Quasi-experiment: comparing group means without random assignment
Comparing cost of care for men and women
Quasi-experiment:
Correlation compares relationship between two measures with no random assignment
Correlation: relationship between two dependent measures (no random assignment)
Men Women
Cost of
Nursing Care
Looking at relationship
between height and weight
Looking at relationship
between age and salary
Weight
Height
Salary
Age

48. Random sampling vs Random assignment
Random assignment of participants into groups:
Any subject had an equal chance of getting assigned to
either condition (related to quasi versus true experiment)
We know this one
Random sampling of participants into experiment:
Each person in the population has an equal chance
of being selected to be in the sample
Population: The entire group of people about whom a researcher wants to learn
Sample: The subgroup of people who actually participate in a research study

49. Within - participant (same as within - subject) & Between - participant (same as between - subject)
Within-participant design: each subject participates in every
level of independent variable (aka repeated measures)
Between-participant design: each subject participates in only
one level of independent variable
Within or between participant design?
Music make plants grow?
Animals make funnier commercials?
Effect of sleep on memory ability

50. Thank you!
See you next time!!