Chapter 3 Notes: The Normal Distributions
Chapter 3: The Normal Distributions
Overview of Chapter Topics
- Density curves
- Describing density curves
- Normal distributions
- The 68–95–99.7 rule
- The standard Normal distribution
- Finding Normal proportions
- Using the standard Normal table
- Finding a value given a proportion
Density Curves
- Definition: A density curve models the overall pattern of a distribution, summarizing a large number of observations.
- Always lies on or above the horizontal axis.
- Total area under the curve equals 1.
- Steps for Exploring a Distribution:
- Always plot your data (e.g., histogram, stemplot).
- Observe the overall pattern of shape, center, and variability (look for outliers).
- Compute a numerical summary for center and variability.
Example of Density Curves
- Vocabulary scores of 947 seventh-graders represented in a histogram.
- A smooth curve fitting the histogram approximates the observed data pattern.
- Example area under the curve for a score of 6.0 is 0.293, indicating the proportion of scores at or below this value.
Characteristics of Density Curves
- Median: The point that divides the area into two equal halves under the curve.
- Mean: The “balance point” of the curve; both mean and median coincide in symmetric distributions but may differ in skewed distributions.
- Mean (μ) and standard deviation (σ) for actual observations vs. those represented by the density curve.
Normal Distributions
- Definition: Normal distributions are a specific class of density curves that are symmetric, single-peaked, and bell-shaped.
- A Normal distribution is fully described by its mean (μ) and standard deviation (σ).
- Changing the mean shifts the curve along the horizontal axis.
- The standard deviation affects the spread: larger σ results in a flatter curve.
68-95-99.7 Rule
- Approximately:
- 68% of observations lie within 1 standard deviation (σ) of the mean (μ)
- 95% within 2σ
- 99.7% within 3σ
Standardizing and z-Scores
- Standardizing: Converting a value from an original Normal distribution to a standard Normal distribution (mean = 0, standard deviation = 1).
- Formula: ( z = \frac{x - \mu}{\sigma} ) (where x = original value)
- Example: For women’s height with a mean of 64.1 inches and a standard deviation of 3.7 inches, a height of 70 inches has a z-score of 1.59, indicating it is 1.59 standard deviations above the mean.
The Standard Normal Distribution
- A standard Normal distribution has a mean of 0 and a standard deviation of 1.
- Any Normal distribution can be standardized to this format.
Cumulative Proportions
- Definition: The cumulative proportion for a value represents the proportion of observations less than or equal to that value.
Example Calculation
- For SAT scores with a mean of 1059 and standard deviation of 210, to find the proportion of students scoring 980 or above, standardize the score:
- ( z = \frac{980 - 1059}{210} = -0.38 )
- Use standard normal tables or software to calculate the area under the curve.
Normal Calculations
- Using Standard Normal Table:
- Define observed variable x (set plot).
- Standardize x to a z-score.
- Refer to Table A for area calculations.
- Finding Scores Given a Proportion:
- To find a score for the top 10%, find the z-score corresponding to 0.90 cumulative proportion (1.28 from tables), then unstandardize:
- Formula: ( x = \mu + z \cdot \sigma )
- For SAT, ( x = 531 + 1.28 \cdot 104 \approx 664.12 )
Conclusion
- Mastery of the normal distribution involves understanding density curves, z-scores, and how to manipulate the mean and standard deviation to calculate proportions and values within a dataset.