CS 1173 Computation for Scientists and Engineers Lab 5
Gender differences in infants

Objectives: Use normplot to assess normality of a dataset. Apply the two-sample t-test to assess whether the means of the underlying populations of two samples are likely to be the same. Practice manipulating rows and columns of an array. Use overlaid scatter plots to assess similarity of distributions. Calculate the correlation between variables to measure similarity. Interpret data and draw conclusions.

Objectives:

Hand-in Requirements:

• The lab5Script.m script containing your work for all parts.
• An MS Word file containing the required bullet points.

Overview

This laboratory looks at head size, length, and weight of female and male infants in the United States. The underlying question is whether male and female infants have significantly different characteristics at birth. The data corresponds to hypothetical individuals and was derived from the standard pediatrician growth charts available from the National Center for Health Statistics of the Center for Disease Control at http://www.cdc.gov/growthcharts/.

Description of the data

• InfantLengthWeightMale.csv contains length-weight data for 1000 male infants. The first column gives infant length in centimeters (cm), and the second column gives the weight of the same infant in kilograms(kg).
• InfantLengthWeightFemale.csv contains length-weight data for 1000 female infants. The first column gives infant length in centimeters (cm), and the second column gives the weight of the same infant in kilograms(kg).
• InfantHeadSizes.csv contains two columns of data. The first column contains the head circumference for 800 male infants in centimeters (cm). The second column gives the head circumference for 800 female infants in centimeters (cm).

Part I: Input the data

• Create a directory called lab5 in Z:\working\MATLAB and ownload the data files to this directory.
• Create a new M-file called lab5Script.m that will contain your commands for this laboratory.
• Put the commands for each part in a separate cell. (You may further subdivide sections of the code into additional cells.)
• Read the data sets into MATLAB.
• Create a variable that contains an array of the male and female lengths (in column 1 and column 2, respectively).
• Create another variable that contains an array of the male and female weights (in column 1 and column 2, respectively).
• Check your workspace to be sure that you have created the variables you expected. Look at their respective sizes and shapes.

Part II: Test for normality using norm plots and the Lilliefors test for normality

• Create a new figure containing the norm plot of infant head sizes (both male and female). (Use xlabel, title, and legend to label the graph rather than editing with the MATLAB plottools.)
• Create a similar figure showing the norm plots of male and female infant weights.
• Create a similar figure showing the norm plots of male and female infant lengths.
• Apply the lillietest to each of the 6 variables to test normality.

Part III: Compare length versus weight for male and female infants

• Create a scatter plot of length versus weight for male infants and length versus weight for female infants. Display males and females in different colors on the same plot. (Use xlabel, ylabel, title, and legend to label the graph rather than editing with the MATLAB plottools.)
• Calculate and output the correlation between male infant length and male infant weight.
• Calculate and output the correlation between female infant length and female infant weight.

Part IV: Are male and female infants different size at birth?

• Apply the two-sample t-test (ttest2) to test whether the mean head sizes for male and female infants are likely to be different.
• Output the hypothesis test value (h), the p-value, and the confidence interval values that were returned by ttest2. Also output the mean head size for male and for female infants.
• Do a similar comparison for infant length.
• Do a similar comparison for infant weight.

Part V: Develop bullet points to summarize the results

• What is the context? The first set should contain five bullet points (who, where, what, when and why)
• Are the data sets normally distributed? The second set should contain three bullet points (indicating whether each of the three data sets is likely to be normally distributed and the reason)
• What is the relationship between length and weight for males and females? Is there a gender difference? The third set of bullet points should contain:
  • At least one observation about the relationship between length and height for male infants
  • At least one observation about the relationship between length and height for female infants
  • At least one observation comparing results for the two genders
  Ideally, you should try to make more than one observation in each category.
• Do infant size characteristics show a dependence on gender? The fourth set of bullet points should contain three bullet points:
  • Description of the difference in average length between male and female infants, with an indication of the basis for comparison.
  • Description of the difference in average weight between male and female infants, with an indication of the basis for comparison.
  • Description of the difference in average head size between male and female infants, with an indication of the basis for comparison.

Grading rubric (point values)

This laboratory was written by Kay A. Robbins of the University of Texas at San Antonio and last modified on 09-Nov-2011. Please contact krobbins@cs.utsa.edu with comments or suggestions. Image is an icon of the Birth to 36 months: Boys Head circumference-for-age and Weight-for-length percentiles available from the Centers for Disease Control.