Detailed Study Notes on Test Management, Grading, Z Scores, and Hypothesis Testing

Test Scores and Grading Expectations

• Announcement regarding test scores:

  • TAs requested to have grading completed by Sunday night.

  • Grades will be discussed the following Monday.

  • Encouragement to keep expectations in check; possibility of pleasant surprise if grades are available earlier on Friday.

• Office hours to review exams:

  • Open office hours will be scheduled for students to come in without an appointment to review their scores.

  • Exact timing to be announced on Canvas.

• Queries regarding test one grades:

  • Instructor acknowledges student anxiety and assures effort is being made on grading.

  • More details on grading will be provided soon.

Schedule Changes

• Class recording adjustments:

  • Instructor decided to hold in-person classes on certain dates: the 27th and the second, instead of recording them.

  • This adjustment is made to enhance student interaction and question-asking opportunities.

• Review day for class on March 2:

  • Expected to be a review day.

  • Possibility of canceling class on the 2nd if students are ahead; may allow for lab review during class.

• Test scheduling flexibility:

  • Original date for test was Friday, March 6, before spring break.

  • Students can now choose to take the test on March 4 or stick with the original date of March 6.

  • If students opt to take the test on March 6, they must provide a valid excuse for any absence on that day to retake the exam.

• Accommodations for students with needs:

  • Those with accommodations can take the test on the day before, the day of, or the day after the scheduled test to provide flexibility due to logistical challenges (traveling to West Campus).

Test Format and Content for Test Two

• Changes in test structure:

  • Test two will not include true or false questions, which the instructor found problematic.

  • Comparison to test one:

  • Test one had:

    • 20 multiple choice questions (2 points each)

    • 10 short answer questions (definitional and practice problems)

  • For test two:

    • 25 multiple choice questions (adding 5 questions for 10 additional points)

    • 5 short answer questions (all based on previously offered practice problems).

• Emphasis on practice:

  • Encouragement for students to practice problems as it will help with speed during the exam.

  • Review sessions will be provided ahead of the test.

Cumulative Nature of Test Material

• Clarification on cumulative content:

  • The test is not cumulative in terms of regressing to previous material (e.g., standard deviation calculations).

  • However, understanding concepts from previous chapters is essential as they build on each other.

  • Prevalent formulas will focus on new material learned in the current chapters.

• Assessment method for student performance:

  • Instructor will monitor completion rates and correctness of test submissions for future adjustment.

  • Goal is to ensure fair treatment in grading.

Z Scores and Percentile Calculation

• Introduction of z scores:

  • Z scores allow for standardization of test scores across different distributions.

  • The conversion from a z score back to a raw score is possible given the mean and standard deviation of the distribution.

• Example calculation:

  • For the SAT's 85th percentile with a mean of 1050 and standard deviation of 210:

  • Determine the z score for the 85th percentile.

  • Calculation steps include finding the area on the z chart corresponding to 35% (as 50% is the mean).

  • The found z score (1.04) can be plugged into the conversion formula:
    $z = \frac{x - \mu}{\sigma}$

  • Re-arranged becomes:
    $x = z \cdot \sigma + \mu$

  • With the final calculation resulting in:
    $x = 1.04 \cdot 210 + 1050 = 1268.40$

  • Conclusion that a score of 1268.40 or higher qualifies for the 85th percentile.

Implementation of Null Hypothesis Testing

• Overview of null hypothesis significance testing:

  • The null hypothesis indicates that there is no effect or difference (e.g., medication has no effect on depression rates).

  • Research hypothesis posits there is a significant effect or difference (e.g., medication decreases depression).

• Visual representation:

  • Normal distribution curves illustrate expectations under null hypothesis.

  • Sample data can help understand if the results are statistically significant.

• Explanation of statistical significance:

  • A 5% significance level is standard; conclusions are only drawn when data falls into this rare category.

  • Comparison made to legal standards, like reasonable doubt in a jury trial.

• Practical examples:

  • Assessing the fairness of a coin through repeated flips and determining at what point the coin appears unfair.

  • The probability thresholds assist in decision-making regarding rejecting the null hypothesis based on observed data (e.g., excess heads or tails seen in flips).

Summary of Key Concepts

• Understanding the mechanisms of z scores and their connection to standardization.

• Importance of hypothesis testing within statistical analysis frameworks.

• Revising core statistical concepts to maintain coherence throughout statistical reasoning and implementation.