Biostatistics (HS167) Lab Manual

Introduction

The biostatistics lab activity is an important part of the biostatistics course. Complete all lab work exactly
as specified in this manual. Keep a record of your lab work in your Procedure Notebook. All lab work
must be completed each week.

Homework exercises are separate, and do not go into the Procedure Notebook.

The premise of the lab is this: You take a simple random sample from a population (Appendix 1) and then
analyze the data in your sample. You will see that learning about the population from data in any given
sample is not as easy as you might think. To help gauge the information in your sample against “what
actually is,” here is some information about the 600 individuals in the population:

Codes and Parameters

# Variable Name Variable Label (Dots represent missing data)
1 ID Identification number

2 AGE Age in years µ = 29.505, σ = 13.59 min = 1, max = 65

3 SEX Gender F (26.5%), M (66.7%), . (6.8%)

4 HIV HIV test results Y (76.8%), N (23.2%), . (0.0%)

5 KAPOSISA Kaposi sarcoma Y (52.8%), N (47.2%), . (0.0%)

6 REPORTDA Report date µ = 7/15/89, min = 01/02/89, max = 02/05/90

7 OPPORTUN Opportunistic infection. Y (60.2%), N (35.3%), . (4.5%)

8 SBP1 Systolic BP No. 1 µ = 120.13, σ = 18.53

9 SBP2 Systolic BP No. 2 µ = 119.95, σ = 19.07

During the first class you must obtain a College of Applied Sciences and Arts (CASA) computer account.
Keep your CASA computer ID and password in a safe place. You are responsible for keeping your
computer account functional. If you experience difficulties with your computer account, contact the
CASA technical staff through the Dean's office (MH433) or by Email ([email protected]). You are
also responsible for knowing how to use Windows computers on local area networks. If you do not know
how to use Windows computers, please consider taking HPrf101 before taking this course.

Page 1 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

Lab 1 (Measurement and Sampling)

Purpose: To select a simple random sample from the population and enter these data into SPSS.

1. Random Numbers: You want to select a simple random sample from the population census
listed as Appendix 1 in this manual. The population consists of 600 individuals, many of whom are
HIV positive. The first step in this process is to generate 10 random numbers between 1 and 600.
Start your Web browser and go to the http://www.openepi.com/ website. In the left-hand panel
select Random Numbers. Generate 10 random integers between 1 and 600 and write these
down for future reference.

2. Data: Identify the 10 people in the population with IDs that match your random numbers. By
hand, list these data as a data table in your Procedure Notebook. The data table should list
variable names across its top and have 10 rows (one for each observation) and 8 columns (one
for each variable).

3. Data Entry: You are now going to enter your data into an SPSS .sav file.*
a. After starting SPSS, select “Type data into file.” Click on the Variable View tab
at the bottom of your screen.
b. Type each variable’s name in the column labeled NAME. Use variable names ID, AGE,
SEX, HIV, KAPOSISA, REPORTDA, OPPORTUN, SBP1, and SBP2.
c. In the column labeled TYPE, use “numeric” for scale variables (ID, AGE, SBP1,
SBP2), “String” for nominal variables (SEX, HIV, KAPOSISA, OPPORTUN), and
“Date” for the date variable (REPORTDA).
d. In the column labeled MEASURE, identify variables as scale, ordinal, of nominal, as
appropriate.
e. You may leave blank the columns for variable Width, Decimals, Label, Value, etc..
f. Click the Data Table tab at the bottom of the screen, and then carefully enter your
data into the table.
g. After you’ve entered your data into the data table, save these as LnameF10.sav (e.g.,
GerstmanB10.sav). Store your data in your home directory or on a floppy disk.
Otherwise it will be gone the next time you sign onto the computer!
h. Print the data table (File > Print) and tape it into your Procedure Notebook.

4. Homework: Complete the homework exercises for Unit 1 assigned in class. These are
completed on separate sheets of paper and hand them in for grading at the beginning of the next
lecture. Do NOT put these exercises in your Procedure notebook. See the syllabus for policies
regarding homework assignments.

*
The .sav file format is native to SPSS. This means they are permanent data files that can be
used only by the SPSS software program. You can purchase a student version of SPSS software at the
campus bookstore, as specified on the syllabus.

Page 2 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

Lab 2 (Stem-and-Leaf Plots and Frequency Tables)

Purpose: To explore the AGE data in your sample with a stem-and-leaf plot and frequency table.

1. Stem-and-leaf plot: Construct a stem-and-leaf plot of the AGE data in the sample you selected
last week. (If necessary, review how to construct a stem-and-leaf plot via your lecture notes or
via StatPrimer.) Make certain you label your axis and include an axis-multiplier. Describe the
shape, location, and spread of the distribution.

2. Computerized Analysis: Start SPSS. Open the file which contains your data
(LnameF10.sav) and click Analyze > Descriptive Statistics > Explore.
Place the AGE variable in the Dependent List and click OK.After the program runs, go to
the OUTPUT window and navigate to the Stem-and-leaf plot. How does this plot
compare with the one you constructed by hand? There is more than one way to draw a stem-
and-leaf plot (e.g., using double valued stems), so a difference does NOT necessarily indicate that
your hand-drawn plot has an error.

3. Frequency table: Create a frequency table of the AGE data in your sample with age grouped
into 10-year class intervals. Include columns for frequency counts, relative frequency, and
cumulative frequency.

4. Frequency table with SPSS

a. Click the Variable View tab (toward the bottom of the screen) and create a new
variable (variable 9) named AGEGRP. Make this a numerical variable with width 8 and 0
decimals.
b. In the column called “Label,” enter “Age Group” to give the variable a descriptive label.
c. Click the Data View Tab at the bottom of the screen and classify each age with the
following codes: 1 = 0–9 years, 2 = 10–19 years, 3 = 20–29 years, 4 = 30–39 years, 5 =
40–49 years, 6 = 50–59 years, and 7 = 60–69 years.
d. Click Analyze > Descriptive Statistics > Frequencies, select the
AGEGRP variable, and click OK.
e. Go to the Output Window and navigate to the frequency table for AGEGRP. View
the frequency table compiled by SPSS. How does this frequency table compare with the
one you prepared by hand?

5. Optional: Recoding the data with SPSS. To have SPSS classify data into intervals, click
Transform > Recode > Into Different Variable. You will then be presented
with a series of dialogue boxes. Select AGE as your input variable and AGEGRP2 as your output
variable. Follow the screen prompts to assign codes to each range. Experiment with various
programming options.

6. Complete the homework exercises assigned in class. As always, do these on separate sheets of
paper and hand them in for grading just before the next lecture begins.

Page 3 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

Lab 3 (Summary Statistics)

Purpose: To calculate and interpret summary statistics for the AGE data in your sample.

1. Sample mean and standard deviation: By hand, calculate the mean and standard deviation of
the AGE data in your sample. Report statistics using guidelines from the Publication Manual of
the American Psychological Association. Means and standard deviations should be reported
with 2 decimals above that of the initial data. In addition, make note of the units of measure and
the sample size. For example, the mean age in my sample is 29.00 years (standard deviation =
15.40 years, n = 10).

2. SPSS descriptive statistics: Start SPSS and open your data file LnameFname10.sav. Click
Analyze > Descriptive Statistics > Descriptives and select the AGE
variable for analysis. Compare the mean and standard deviation reported by SPSS to your hand
calculations. If results differ, track down the error and make corrections.

3. 5-point summary & boxplot: Determine the 5-point summary for your AGE data. (Review
StatPrimer p. 3.5, if necessary.) Draw a boxplot of your AGE data. Are there any outside values
in your data set? In plain English, describe the distribution’s shape, location, spread.

4. SPSS exploratory analysis

a. Click Analyze > Descriptive Statistics > Explore.
b. Place the AGE variable into the Dependent List.
c. Click the Statistics buttons, check the percentiles box, and click OK.
d. Go to the Output Window and navigate to the Percentiles section of the outputs.
This section reports quartiles using two methods of calculation. Our method of
corresponds to Tukey’s Hinges and NOT to the Weighted Average percentiles! Ignore
the weighted average percentiles. Make certain your quartiles match Tukey’s hinges.
e. Navigate to the output region with the boxplot and compare the boxplot you drew by hand
with the boxplot created by SPSS. Are they similar?

5. Complete the homework exercises assigned in class.

Page 4 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

Lab 4 (Probability)

Purpose: To calculate and interpret binomial and normal probabilities.

1. A Binomial Problem

a. Probability distribution: Suppose a treatment is successful 25% of the time. The treatment is
used in 3 patients. Using the binomial formula learned in class (formula 4.3 in StatPrimer),
calculate the probability of seeing 0 of three positive responses. Then calculate the probability of
seeing 1 response, 2 responses, and 3 responses. These probabilities comprise the probability
distribution X~b(n = 3, p = .25).

Pr(X = 0) = _____
Pr(X = 1) = _____
Pr(X = 2) = _____
Pr(X = 3) = _____

b. Probability calculator: Log on to your computer. Click the Windows Start Bar and select
StaTable for Windows. (You may download this program from www.cytel.com/statable or use
the Web version of this program.)
i. Click Distributions > Discrete > Binomial.
ii. Fill in the field labeled “P” (the field with the dice) with .25 (since p = .25).
iii. Fill in the field labeled “number of trials N” with 3 (since n = 3).
iv. Fill in the field labeled i with the number of successes (0, 1, 2, and 3, respectively.
v. Probabilities are reported in the field labeled “Pr(i)”. These probabilities should match the
probabilities calculated by hand. If they do not match, track down your error and make
corrections.

c. Draw a probability histogram of your binomial distribution with the “# of successes” on the X
axis and “probability” on the Y axis. The area under each histogram bar corresponds to the
probability of each outcome.

d. The cumulative probability of an event is the probability of seeing exactly that number or less.
Calculate the following cumulative probabilities:

Pr(X # 0) = Pr(X = 0) = ______

Pr(X # 1) = Pr(X = 0) + Pr(X = 1) = ______
Pr(X # 2) = Pr(X = 0) + Pr(X = 1) + Pr(X = 2) = ______
Pr(X # 3) = Pr(X = 0) + Pr(X = 1) + Pr(X = 2) + Pr(X = 3) = ______

e. Return to StaTable for Windows and notice that cumulative probabilities are reported in the field
labeled “LEFT.” Make certain your hand-calculated cumulative probabilities match StaTable’s
“LEFT” values. If a discrepancy exists, track it down and make corrections.

f. On the histogram you created in part c, shade the bars corresponding to Pr(X = 0) and Pr(X =
1). This represents Pr(X # 1). Notice that this is the area in the LEFT TAIL of the distribution.

Page 5 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

 This lab continues on the next page.

Page 6 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

2. Normal probability problem

a. Z table: Go to the StatPrimer website (www.sjsu.edu/faculty/gerstman/StatPrimer). Scroll down

to the bottom of this page and click on the link to the Z table. Print this table (File > Print)
and place it in your Procedure Notebook.

b. Using the Z table

i. Draw a normal probability curve. At the points of inflection on the curve, mark !1 and
+1, respectively. Then, mark the !2 and +2 landmarks.
ii. Use the z table to determine the area under the curve to the left of +1. This is Pr(Z < 1).
iii. Use the z table to determine the area under the curve to the left of +2. This is Pr(Z < 2).
iv. Using the symmetry of the curve, determine the area under the curve to the left of !1.
This is Pr(Z < !1).
v. Using the symmetry of the curve, determine the area under the curve to the left of !2.
This is Pr(Z < !2).

c. Return to the StaTable computer program. Click Distributions > Continuous >
Normal. Use this program to determine Pr(Z < 1), Pr(Z < 2), Pr(Z < !1), and Pr(Z < !2).
These will be reported as areas in the left tail of the distribution.

d. Z percentiles: Let zp denote a z percentile with a left tail area of p. For example, z.50 = 0 since a
z of 0 has a left tail area of .50. Determine each of the following z percentiles:

z.8413 = _____ z.9772 = _____ z.1587 = _____ z.0228 = _____

80
e. The distribution of SPB1 in the population: Using Internet
Explorer, go to www.sjsu.edu/faculty/gerstman/datasets. 60

Right click populati.sav and save it to your Home 40

directory. Open this file from your home directory and click
Graphs > Histogram. Select the SBP1 variable, check 20 Std. Dev = 18.53
Mean = 120.1
the box that says Display Normal curve, and click OK. 0 N = 600.00

(The figure is displayed to the right.) This variable is

65
75
85
95
10
11 .0
12 .0
13 .0
14 .0
15 .0
16 .0
17 .0
.0
.0
.0
.0
5
5
5
5
5
5
5
5.
0
approximately normally distributed with a mean of 120.13 and SBP1

standard deviation of 18.53. We use the notation X~N(µ = Fig. Distribution of SBP1 in the population.
120.13, σ = 18.53) as shorthand to describe this variable.

f. Standardization: Values are standardized by subtracting the distribution’s mean and dividing by
the standard deviation. For example, of 83 for the variable SBP1 has a standard score of Z =
83 − 120 .13
= −2 .00 . This places a value in the left tail of distribution 2 standard deviations
18.53
below average. Since, Pr(Z < !2.00) = .0228, we expect 2.28% of the values in this distribution to
lie below 83. Determine what percentage of values that actually lies below 83 by opening
populati.sav and clicking Analyze > Descriptive Statistics >
Frequencies and creating a frequency table for the SBP1 variable. Was the prediction based
on the normal distribution accurate? How well does the model fit the data?

3. Complete the homework exercises assigned in class.

Page 7 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

Lab 5 (Confidence Interval for a Mean)

Purposes: To learn about distributions of sample means and confidence intervals for means.

200
1. Age distribution in the population: The AGE distribution in
populati.sav is displayed in the figure to the right. This distribution
on has µ = 29.5 and σ = 13.59. What is the mean AGE in your sample?
Why is this mean different from the mean in the population?
100

Parameters:
2. A sampling experiment: The lab instructor will collect each student’s

Frequency
sample. The instructor will then add these sample means to Std. Dev = 13.59
Mean = 29.5
SampleMeans.sav as the variable The variable MEANAGE. This file
0 N = 600.00
will be stored on the LAN in J:\Gerstman. When notified, make a

0.
5.
10
15.0
20 0
25 0
30.0
35 0
40 0
45 0
50.0
55 0
60 0
65 0
0
0

.
.

.
.
.

.
.
.
.0
copy of SampleMeans.sav and place it in your Home (H:) AGE (years)
directory. Open your copy of SampleMeans.sav in SPSS and Fig: Distribution of AGE in the population.
construct a histogram of the variable MEANAGE (Graphs >
Histogram). Check the box that says Display normal curve before clicking OK. The
histogram you just created is a nascent sampling distribution of means. Describe the shape, location,
and spread of this nascent sampling distribution. Is it more normal than the AGE variable in the
population (Fig. above)? What is the mean of MEANAGE? What is the standard deviation of
MEANAGE? Explain how these statistics relate to the central limit theorem and law of averages.

3. Confidence interval for µAGE, σ known: Using the mean in your sample and the standard deviation
in the population (σ = 13.59), calculate a 95% confidence interval for the population mean age. (See
StatPrimer p. 5.6, if necessary.) Did your confidence interval capture the value of population mean?
In the long run, what percentage of 95% confidence intervals will fail to capture µ?

4. t Table and t percentiles: Go the StatPrimer website, scroll to the bottom of the page, and click on
the link to the t table . Print this table (File > Print) and place it in your Procedure Notebook.
Let tdf,p denote a t percentile with df degrees of freedom with a left tail area of p. Then, using the t
table, determine the percentiles listed below.

t9,.90 = _____ t9,.95 = _____ t9,.975 = _____ t9,.995 = _____

After determining the above t percentiles, check your findings with StaTable for Windows
(Distributions > Continuous > Student’s t).

5. Confidence interval for µAGE, σ estimated: Use your sample mean age and sample standard
deviation to calculate 95% confidence intervals for µ. (See StatPrimer p. 7, if necessary.) Then, open
your data set (LnameF10.sav) in SPSS and check your confidence interval calculations with
Analyze > Descriptive Statistics > Explore (select the AGE variable). The
confidence interval is reported in the output area labeled “Descriptives.”

6. Complete the homework exercises assigned in class.

Page 8 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

Lab 6 (Testing a Mean)

Purpose: To test a sample mean against a hypothesized “null value.”

Background: The mean AGE in the population (µAGE ) = 29.5 years and the
standard deviation (σAGE ) = 13.59.

1. Sampling distribution: From the central limit theorem, we know that the sampling distribution of
mean AGES based on n = 10 taken from populati.sav will tend toward normality with µ = 29.5
and a standard deviation equal to SEM = 13.59 / /(10) = 4.3. (The sampling distribution of means is a
hypothetical model that does not actually exist in nature.) Thus, x ~N(µ=29.5, σ=4.3) and
approximately 95% of sample means 10 will fall in the interval 29.5±(1.96)(4.3) = 29.5 ± 8.4 ≅ (21,
38).

95%

What percentage of sample means will fall within one (1) standard error of the mean?

2. Sampling experiment: Retrieve the file SampleMeans.sav from your home directory. (You
used this file during the last lab.) Using SPSS, construct a frequency table of the variable MEANAGES
(Analyze > Descriptive Statistics > Freq). What percentage of sample means
were actually greater than 38? What percentage were actually less than 21?

3. Z test: Assume you do not know the mean age of the population but do know the standard deviation
(σ = 4.3). Suppose an investigator hypothesizes that the mean age of the population is 32. (The
investigator is wrong.) Perform a two-tailed z test of H0 : µ= 32. Let α = .05. State all 4 statistical
hypothesis testing steps as discussed in class. (See StatPrimer pp. 6.4 & 6.5, if needed.) If you
retained the null hypothesis (which is likely), what type of error have your made, a type I error or type
II error?

4. T test: Now assume σ is not known. Thus, you will estimate σ based on the standard deviation (s) in
your sample. Conduct a two-tailed t test of H0 : µ = 32. Let α = .05. (See StatPrimer p. 6.6, if
needed).

5. Check your t test calculations with SPSS: Open your data set (LnameFname10.SAV) with
SPSS and click Analyze > Compare Means > One Sample T Test. Place AGE in the
field labeled test variable and enter “32” in the field labeled test value (since you are testing H0 : µAGE
= 32.) Switch to the output window and review the test statistics calculated by SPSS. These results
should match the results you calculated in step 4.

Page 9 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

6. Complete the homework exercises assigned in class.

Page 10 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

Lab 7 (Paired Samples and Their Differences)

Purposes: To describe differences in paired samples, calculate a confidence interval a paired mean
difference, and test a paired difference for significance.

The mean difference of SBP1 and SBP2 in the population (µd ) = 0.18.

1. Paired samples: Your data set includes the variables SBP1 and SBP2. These represent paired
systolic blood pressure measurements (Hg mm) in individuals. Open your data set
(LnameFname10.sav) in SPSS and calculate descriptive statistics for these two variables
(Analyze > Descriptive Statistics). How do these two measurements compare? (i.e.,
Compare their means and standard deviations as measures of central location and spread.)

2. Difference variable DELTA: By hand, calculate differences for each matched-pair (DELTA =
SBP1 - SBP2). After differences are calculated, construct a stem-and-leaf plot of DELTA.
Describe the shape, location, and spread of this distribution.

3. Mean and standard deviation of DELTA: By hand, calculate the mean and standard deviation of
DELTA.

4. SPSS: Go to the Data View window in SPSS. Click Tranform > Compute. Let the target
variable be DELTA and the numeric expression SBP1 - SBP2 and then click OK. in your Data
View window, SPSS will create the variable DELTA with calculated differences. The click
Analyze > Descriptive Statistics > Explore and place DELTA in the Dependent
List. Go to the output window and review the descriptive statistics, stem-and-leaf plot, and boxplot
calculated by SPSS.

5. 95% confidence interval for µdelta: Calculate a 95% confidence interval for the mean difference.
(Review StatPrimer p. 7.3, if necessary.) Interpret your results. Did your confidence interval capture
µd ? SPSS calculated this 95% confidence in the prior step. Compare your confidence interval to the
one calculated by SPSS.

6. Statistical hypothesis testTest H0 : µdelta = 0. Let α = .01. List all hypothesis testing steps. Interpret
your results. (You may check your calculations with SPSS by clicking Analyze > Compare
Means > One Sample T Test, Test variable = DELTA, Test value = 0.)

7. Power: Determine the power of the above test to detect a difference of 5 mm Hg while assume σd =
5. Let α = .05 (two-sided) by using the method described on StatPrimer p. 7.5. Was the power of
the test adequate?

8. Complete the homework assigned in class.

Page 11 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

Lab 8 (Independent Sample and their Differences)

Purposes: To describe independent samples, estimate a mean difference with 95% confidence, and
conduct an independent t test.

The mean difference in AGE in HIV+ and HIV− people in the population (µ1 − µ2 ) = !0.64 years.

1. New data file: Retrieve the file GerstmanB10.sav from either the LAN or WWW (see lab
instructor if assistance is needed). Store your copy of this file in your Home directory. Open your
copy of this data file in SPSS and make note of AGE values for the HIV+ and HIV! people in this
sample.

2. Side-by-side boxplot: Determine 5-point summaries of the HIV+ people (n 1 = 7) and HIV! people
(n 2 = 3) in the sample. Then, construct a side-by-side boxplot of these distributions. Do the
distributions overlap? How do the medians compare?

3. SPSS: Open GerstmanB10.sav in SPSS and click Analyze > Descriptive

Statistics > Explore. Put the variable AGE in the Dependent List and HIV in the Factor
list. Go to the output window and navigate to the boxplot. How does this boxplot compare with the
one you produced by hand?

4. Mean and standard deviation: Calculate the mean and standard deviation of each group.

5. Confidence interval for independent mean difference: Calculate the pooled estimate of variance
(StatPrimer formula 8.1), standard error of the mean difference (formula 8.2), and 95% confidence
limits for µ1 − µ2 (formula 8.3). Interpret your confidence interval. Did your confidence interval
capture the true mean difference of !0.64 years?

6. Statistical hypothesis test: Test H0 : µ1 − µ2 = 0. (Review StatPrimer p. 8.4, if necessary.) List all
hypothesis testing steps. Were you able to reject the null hypothesis? (Probably not.) Does this imply
the null hypothesis is correct? (Absolutely not!) Did you make a type I or type II error?

7. SPSS: Check your calculations with SPSS by clicking Analyze > Compare Means >
Independent Samples T Test. The Test Variable is AGE and the Grouping Variable is
HIV. You must use the Define Groups button to tell SPSS that Groups 1 is coded “Y” and
Group 2 is coded “N.” After the program runs, go to the output window and navigate to the region
labeled “Independent Samples Test.” The first row of the output table (labeled “Equal Variances
Assumed”) contains confidence interval and test statistics. These should match the statistics you
calculated by hand in part 5 and 6, respectively.
2
16 ⋅ σ
8. Sample size requirements for a future study: Use the formula n = 2 + 1 to calculate the
∆
sample size requirements needed to detect a mean difference of 10 with 80% power. Assume σ2 =
223.37.

Page 12 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

9. Complete the homework exercises assigned in class.

Page 13 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

Lab 9 (Inference About a Proportion)

Purposes: To estimate the prevalence of HIV in the population and to calculate the sample size needed to
estimate this prevalence with a margin of error of 0.1 (10%).

The prevalence of HIV in the population = 468 / 600 = .768.

. ) 2 pq
(196
1. Sample size requirements: Use formula n = to determine the sample size needed to
d2
estimate the prevalence of HIV in the population with a margin of error of 0.10 (10%). Since you do
not have a good estimate for p, assume p = .5 for now. This will ensure the formula derives a more-
than-adequate sample size.

2. New data set: Go to www.sjsu.edu/faculty/gerstman/datasets and download

GerstmanSampleBig.sav to your home (H:) directory. This file has 96 observations with the
HIV variable coded: 1 = HIV+ and 2 = HIV-. Open your copy of this file with SPSS and determine
the prevalence of HIV in the sample. Then, use the “npq rule” to see whether the normal
approximation can be used with data set.

3. 95% confidence interval for the prevalence: Calculate a 95% confidence interval for parameter
p using the data in GerstmanSampleBig.sav and formula 9.3. Interpret your results. Did the
confidence interval capture the true prevalence?

4. Statistical hypothesis test: An investigator wants to test whether the prevalence of HIV in the
population is greater than 50% . Thus, H0 : p = .5, one-sided. Using data in
GerstmanSampleBig.sav, perform this test. (See StatPrimer p. 9.4, if necessary.) Let α =
.01. After you’ve completed your calculations, use SPSS to check your work by clicking Analyze
> Nonparametric > Binomial: Test variable = HIV, Test proportion =
.50.

5. Complete the homework exercises assigned in class.

Page 14 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)

Lab 10 (Cross Tabulated Counts)

Purposes: To estimate the prevalence difference of HIV in men and women in the population; to test
whether there is a significant association between SEX and HIV.

The prevalence of HIV in men in the population (p 1 ) is .230

The prevalence of HIV in women in the population (p 2 ) is .220
The prevalence difference = .230 ! .220 = .010

1. Cross-tabulation: Open GerstmanSampleBig.sav and click Analyze > Descriptive

Statistics > CrossTabs. Select SEX as your row variable and HIV as your column
variable. Make note of the cross-tabulation.

2. Prevalence difference: Calculate the prevalence of HIV in men ( p$1 ) and women ( p$ 2 ). Calculate
the prevalence difference. Why does this prevalence difference differ from the prevalence difference
in the population?

3. 95% confidence interval for prevalence difference: Calculate a 95% confidence interval for the
prevalence difference. (See StatPrimer p. 10.2, if necessary.)

4. Expected frequencies under H00 : Create an expected frequency table for the above cross-
tabulated counts. Are any expected frequencies less than 5? Can a chi-square test be used with these
data?

5. Chi-square table: Go the StatPrimer page on the web and scroll toward its bottom. Click on the link
for the chi-square table, print a copy of this table, and place it in your Procedure Notebook.

6. Chi-square percentiles: Let χ 2 df,p represent a chi-square percentile of p with df degrees of

freedom. Determine the following chi-square percentiles:

χ 2 1,.90 = ______ χ 2 1,.95 = ______ χ 2 1,.99 = ______

χ 2 2,.95 = ______

7. Chi-square test: Perform a chi-square test of the cross-tabulated data. Let α = .05. List H0 and H1 .
Show all hypothesis testing steps and calculations. Is there a significant association between HIV and
SEX?

8. SPSS chi-square test: After calculating chi-square statistics by hand, compute the test with SPSS.
This is done by clicking Analyze > Descriptive Statistics > CrossTabs. >
Statistics button. Check the chi-square box and click OK (etc.).

9. Complete the homework exercises assigned in class.

Page 15 of C:\DATA\HS167\lab-manual.wpd (Print date: 6/22/03)