Statistics 101L – Laboratory 4
In this lab we will investigate whether we can use a Normal model to describe the
distribution of ACT Composite scores.
Below is a table of Percentile Ranks for Enhanced ACT Composite scores for freshmen
entering ISU in Fall 2000.
ACT
Composite
16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
Percentile
Rank
1 3 5 10 16 24 33 43 53 62 71 77 83 88 92 95 97 99 100
The Percentile Rank is interpreted in the following way; 62% of all ACT Composite
scores are 25 or below. Similarly, 53% of all ACT Composite scores are 24 or below.
From this we know that 9% of all ACT Composite scores are equal to 25.
1. Create a new table indicating the percentage of all ACT Composite scores that are
equal to each of the values 16, 17, … , 34.
2. Construct a histogram based on the information from the table you create in 1. To
do this using JMP first create a data table with two columns. One column is for
the ACT scores and the second column is for the corresponding percentages. Use
Analyze – Distribution and put the ACT scores as the Y, Columns and percentage
as the Freq. You will probably have to use the Axis Settings and the “grabber” to
create a “nice looking” histogram. Include a probability axis on the histogram.
You can also have JMP create a Normal Quantile Plot.
3. Describe the shape of the histogram.
4. Consider a Normal model for the ACT Composite score with μ=24.5 and σ=4.0.
Sketch this model on top of you histogram. If you have used JMP to create the
histogram you can chose to Fit Distribution – Normal. JMP will estimate a mean
and standard deviation but you can change these to 24.5 and 4.0, respectively by
going to the red triangle pull-down menu next to Fitted Normal and choosing Fix
Parameters and entering User Defined Values.
5. Construct a table similar to the percentile rank table above except use the Normal
model with μ=24.5 and σ=4.0 to compute the probability of being less than or
equal to each of the ACT values 16, 17, …, 34. Use the web site
http://davidmlane.com/hyperstat/z_table.html to compute these probabilities.
Rounded the probabilities to 2 decimal places and report them as percentages.
6. How do the percentages computed in 5 compare to the Percentile Ranks in the
table above?
7. Based on your answers above do you think that a Normal model with μ=24.5 and
σ=4.0 is a reasonable model for ACT Composite scores? Explain briefly.
