1) A study was conducted to examine whether students' expectations of their AP test scores improved after taking the test. Surveys were given to students before and after taking the AP English and Government tests. 2) Chi-squared tests found the distributions of predicted scores did not match national averages, both before and after taking the tests. 3) Matched pairs t-tests found no significant difference between students' predicted scores before and after taking the AP English and Government tests. 4) An independent t-test also found no significant difference in the change of predicted scores between the two tests.

1. AP STATISTICS
By Ese Uwhuba & Hayden Hilliard.

2. ABSTRACT
A study on the level of AP readiness and the effect of taking the test on a
students expected score was conducted at Skyline High School. The purpose of
the study was examine whether students expectations about their scores improved
after actually taking the test. In performing this study, we expected an increase or
decrease in expected score based on initial stated level of preparation. That is
students who stated that they felt adequately prepared are expected to predict
either an increase or a constant score after taking the test and conversely. The
study is also concerned with the general AP (predicted) score distribution in
comparison to national standardized levels from a recent testing session. This
aspect of the study will be particularly helpful in determining the effectiveness of
AP classes and teachers at Skyline.
If you have ever wondered: Should I know this? Why does my teacher never teach
me anything? Or The year’s almost over, What have I really learned?
OR Is it strange that I do not know the basis facts about the test?
OR, this was a total Then get ready to see the real
waste of $87 reasons….

3. HOW PREPARED ARE
SKYLINE STUDENTS FOR Question
THEIR AP TESTS?

4. T YPICAL HIGH SCHOOL STUDENT
So much Stresses….How de we do it?

5. SAMPLING TECHNIQUE
 We obtained a list from our teacher of everyone taking the AP
test of ficially. We then assigned each person a number ( 1-83
for English and 1-89 for Government) and used a random
number generator to determine a sample of 30. Each person
selected for the sample was then assigned a number 1-30 and
given a survey to protect their identity. If there was anyone
not taking Government that we originally surveyed for English,
we replaced them with someone of equal merit (at our
discretion) to be surveyed for Government

6. 6
Data: English Predicted Scores Before/After
Y or N Before After Y or N Before After
Y 3 3 Y 5 4
N 3 2 N 3 3
N 4 4 N 3 3
N 5 4 N 3 4 N 3 2
N 3 2 N 3 4 N 3 1
Y 3 4 N 4 3 N 3 3
N 3 3 Y 4 4 Y 4 4
Y 4 3 N 2 2 N 2 3
N 3 3 N 2 2 Y 4 3
Y 3 3 N 4 4
N 3 4 N 3 3
N 3 4 Y 3 3

7. 7
Data: Govt. Predicted Scores Before/After
Y or N Before After Y or N Before After
N 3 3 N 3 4
Y 4 2 Y 3 3
Y 4 4 N 2 3
Y 4 4 Y 5 5
Y 4 4 N 3 3
Y 4 4 Y 3 3
N 4 4 N 3 3
N 3 4 N 3 4
Y 4 4 Y 4 4
N 3 3 N 3 3
Y 4 5 N 3 4
Y 4 4 Y 4 5
Y 4 5 N 3 3
Y 4 3 N 1 3
N 3 3 N 4 4

8. 8
Do students feel prepared for their AP tests: Yes or No
How Prepared do students feel for
the AP English test?
30%
70% Yes
No
How prepared do students feel for
the AP government test?
Yes
No

9. 9
Predicted Scores: Before and After.
Predicted Scores Before the Test Predicted Scores Before Test
20
16
18 14
14 13
16
14 12
Frequency
12 10
10 8
8 6
6
4
4
2 1 1 1
2
0 0
1 2 3 4 5 1 2 3 4 5
AP SCORE
Predicted Test Score After Test Predicted Scores After Test
14 14 13
12 12 11
10 Frequency
10
8 8
6 6
4 4
4
2
2 1
0 0
0
1 2 3 4 5
1 2 3 4 5
APENGLISH APGOVERNMENT

10. 10
HYPOTHESIS TESTING
Chi-Squared test of Goodness of Fit
Matched Pairs T Test
2 Sample T test

11. 11
X 2 Goodness of Fit, English
• Question: Are the proportions of predicted scores equal to
the distribution of the national average?
P1 = Proportion of test
Ho: P1 = 0.428
scores within 1-2
P2 = 0.31 P2 = Proportion of test
scores that equal 3
P3 = 0.262 P3 = Proportion of test
Ha: P1 ≠ P2 ≠ P3 scores within 4-5
α = .05
Assumptions
1. Random Sample
2. Expected cell count at least 5 (cells combined)
Before Test X 2 = 16.817 p ≈ 0 ˂ α Reject Ho
After Test: X 2 = 6.672 p = 0.03557˂ Reject Ho α

12. 12
X 2 Goodness of Fit, Government
• Question: Are the proportions of predicted scores equal to
the distribution of the national average?
P1 = Proportion of test
Ho: P1 = 0.428
scores within 1-2
P2 = 0.31 P2 = Proportion of test
scores that equal 3
P3 = 0.262 P3 = Proportion of test
Ha: P1 ≠ P2 ≠ P3 scores within 4-5
α = .05
Assumptions
1. Random Sample
2. Expected cell count at least 5 (cells combined)
Before Test: X 2 = 17.2104 p 0 Reject Ho
After Test: X 2 = 21.857 p 0 Reject Ho

13. 13
Matched Pairs Test, English
• Question: Is there a significant difference between
students Before & After score predictions?
Ho: μd = 0 μd = μ1 - μ2 = 0
μ1 = Mean of predicted test scores before
Ha: μd ≠ 0 taking the test
Assumptions μ2 = Mean of predicted test scores after taking
the test
1. Random Sample α = .05
2. Samples are paired (before, after)
3. Samples are large n ≥ 30.
t = 0.72177 p = 0.4762 > α
Fail to Reject Ho

14. 14
Matched Pairs Test, Government
• Question: Is there a significant difference between
students Before & After score predictions?
Ho: μd = 0 μd = μ1 - μ2 = 0
μ1 = Mean of predicted test scores before
Ha: μd ≠ 0 taking the test
Assumptions μ2 = Mean of predicted test scores after taking
the test
1. Random Sample α = .05
2. Samples are paired (before, after)
3. Samples are large n ≥ 30.
t = -1.75568 p = .0897 Fail to Reject Ho

15. 15
2 Sample t Test
• Question: Is there a significant difference in the means of
the difference between before and after test scores for the
English and Government Tests?
Ho: 1 - 2 = 0
Ha: 1 2
Assumptions
1. Both samples are Independently selected random
samples (They are random & the results of one sample
do not affect the result of the other)
2. Large Sample size n ≥ 30.
t = -1.887 p = .06412 Fail to Reject Ho