Standard Normal Distribution: Unit Normal Table and Z-Score Calculations
Basic Problem Types for Standard Normal Distribution
- There are two fundamental types of problems encountered when working with the standard normal distribution:
- Type A: Looking up a proportion (p) given a specific z-score.
- Type B: Looking up a z-score given a specific proportion (p).
- These two types represent the inverse of each other; one moves from the axis value to area, while the other moves from area to the axis value.
Fundamental Definitions and Interchangeable Terminology
- Within the context of a normal distribution, several terms are functionally interchangeable as they all refer to the shared area under the density curve:
- Proportion: Expressed as a decimal between 0 and 1.
- Percent: The proportion multiplied by 100 (moving the decimal two places to the right).
- Probability: The likelihood of randomly selecting a specific score or range of scores.
- Area: The physical region shaded under the bell curve.
- Conversion between proportions and percentages is a core skill:
- To convert a proportion to a percent: Move the decimal point two spots to the right.
- To convert a percent to a proportion: Move the decimal point two spots to the left.
- Two tools are required to solve these problems. It is taught that one cannot solve these problems accurately using only one tool; both must be used in conjunction.
- The Unit Normal Table (or z-table):
- Column A: Contains the z-scores.
- Column B: Contains the proportion in the Body (the larger portion).
- Column C: Contains the proportion in the Tail (the smaller portion).
- Column D: Not focused on in this specific session.
- The Drawing:
- A normal distribution curve (bell curve) must be drawn for every problem.
- The center of the distribution is always marked as z=0. This point represent the mean, where the score is zero standard deviations away from the mean.
- Positive z-scores are placed to the right of zero; negative z-scores are placed to the left.
Determining Shading Direction through Keywords
- Keywords within a word problem indicate the direction of shading relative to the z-score. This is critical for identifying whether the objective is a "body" or a "tail."
- Shading to the Right:
- Keywords: "Above," "Greater than," "Top," "To the right."
- Notation: p(z>value).
- Shading to the Left:
- Keywords: "Below," "Less than," "Bottom," "To the left."
- Notation: p(z<value).
- Important caveat: Directional keywords like "top" or "below" refer to the direction of shading, not the sign of the z-score itself.
Understanding the "Body" versus the "Tail"
- Slicing a normal distribution at any point other than the exact center (z=0) creates two unequal sections:
- The Body: The larger portion of the distribution (greater than 0.5 or 50%" ").
- The Tail: The smaller portion of the distribution (less than 0.5 or 50%" ").
- If the distribution is sliced exactly at the middle (z=0), the area is perfectly divided into two halves (0.5 and 0.5), meaning there is no distinct body or tail.
Finding Proportions from Z-Scores (Type A Problems)
- Example 1: Proportion above z=1.3
- Step 1: Draw the curve and mark z=0 at the center.
- Step 2: Place z=1.3 to the right of zero.
- Step 3: Shade to the right (keyword: "above").
- Step 4: Identify the shaded area as the Tail.
- Step 5: Look up z=1.3 in Column A and find the value in Column C.
- Result: Proportion = 0.0968 (9.68%" ").
- Example 2: Proportion below z=1.3
- Step 1: Re-use the drawing for z=1.3, but shade to the left.
- Step 2: Identify the shaded area as the Body.
- Step 3: Look up z=1.3 in Column A and find the value in Column B.
- Result: Proportion = 0.9032 (90.32%" ").
Handling Negative Z-Scores and the Principle of Symmetry
- The Unit Normal Table typically only lists positive z-scores because the normal distribution is perfectly symmetrical.
- Example 3: Percent above z=−1
- Step 1: Mark z=−1 to the left of zero.
- Step 2: Shade to the right (keyword: "above").
- Step 3: Identify the shaded area as the Body.
- Step 4: Look up the positive version (z=1) in the table and find the Body column.
- Result: Proportion = 0.8413 (84.13%" ").
- Example 4: Probability of selecting a z-score below z=−1
- Notation: p(z<−1).
- Step 1: Mark z=−1 to the left of center and shade to the left.
- Step 2: Identify the shaded area as the Tail.
- Step 3: Look up z=1 in the table and find the Tail column.
- Result: Proportion = 0.1587 (15.87%" ").
Finding Z-Scores from Proportions or Percentages (Type B Problems)
- In these problems, the area (proportion/percent) is given, and the goal is to find the corresponding z-score on the axis.
- The logic of the drawing changes:
- In Type A, the drawing identifies Body/Tail.
- In Type B, the drawing primarily identifies the sign of the z-score (positive or negative).
- Heuristic for sign determination:
- If you need a tail (p<0.5) and are shading top/above, the z-score is positive.
- If you need a tail (p<0.5) and are shading bottom/below, the z-score is negative.
- If you need a body (p>0.5) and are shading top/above, the z-score is negative.
- If you need a body (p>0.5) and are shading bottom/below, the z-score is positive.
Case Studies for Finding Z-Scores from Area
- Example 5: The top (above) 5.05% of scores
- Step 1: 5.05%=0.0505. This is a Tail (less than 0.5).
- Step 2: Shade the "top" (right side). To get a small tail by shading right, the line must be on the right side of zero (positive z).
- Step 3: Find 0.0505 in Column C (Tail column).
- Step 4: Find the corresponding z-score in Column A.
- Result: z=+1.64.
- Example 6: The bottom (below) 5.05% of scores
- Step 1: 0.0505 is a Tail.
- Step 2: Shade the "bottom" (left side). To get a small tail by shading left, the line must be on the left side of zero (negative z).
- Step 3: Sign is negative.
- Result: z=−1.64.
- Example 7: The top (above) 90% of scores
- Step 1: 90%=0.9000. This is a Body (more than 0.5).
- Step 2: Shade the "top" (right side). To get a massive body by shading right, the cut must start deep in the left side of the distribution (negative z).
- Step 3: Look for 0.9000 in the Body column (Column B).
- Step 4: The values nearest on the table are 0.8997 (z=1.28) and 0.9015 (z=1.29). Both are acceptable approximations.
- Step 5: Apply the negative sign identified in Step 2.
- Result: z=−1.28 or z=−1.29.
Special Cases and Final Reminders
- The 50% Threshold: If the problem asks for the bottom 50%" " or top 50%" ", the z-score is always exactly 0. This is because the mean splits the normal distribution into two equal halves.
- The Importance of the Sign: Missing a negative sign is considered a major error, not a small oversight. A z-score of +1.64 (way above average) is qualitatively different from −1.64 (way below average). Correct identification of the sign proves a correct drawing was made.
- Approximations: Often, the exact proportion requested (like 0.9000) is not on the table. Choosing the closest value available is the standard procedure for these exercises.
- Laboratory Practice: The speaker advises rewatching this specific instructional video while completing the laboratory assignments, as the procedural steps are the core aspect of the record.