SPSS (biometry)

Testing for the presence or absence of a specific disease, medical condition, or illegal drug is common. The results of these tests are never as simple as they appear to be on many TV shows and movies. As patients become more and more critical consumers of medical information from their doctors, they must be aware of the quantitative and statistical reasoning that lurks behind the reported facts and figures. For example, if a medical test is reported as “95% accurate” and you get a “positive” result, what is the chance that you actually have that medical condition?

In this lab, you will explore the results of medical tests. The story that we use is for a made-up disease, but the statistics involved are what medical professionals use to evaluate such tests.

Here are some definitions that you will need:

False positive (FP): when a patient receives an incorrect positive test result for a disease BUT the patient does not have the disease.

False negative (FN): when a patient receives an incorrect negative test result for a disease BUT the patient does have the disease.

True positive (TP): when a patient receives a correct positive result (does have the disease).

True negative (TN): when a patient receives a correct negative test result (does not have the disease).

Sensitivity: the probability that a test correctly produces a positive test result when the patient does have the disease.

Specificity: the probability that a test correctly produces a negative test result when the patient does not have the disease.

A new, faster, cheaper test (we’ll call it CHEAP) has been developed to diagnose a fairly common disease called SpringFeverItis (we’ll call it SFI). The SFI virus causes people to daydream, miss work and school, wear skin-bearing clothing, and spend excessive amount of time in the sun. The manufacturers of this CHEAP test claim that it is 95% effective in detecting when a person has the SFI virus. This new test will be compared to the “gold standard” test (which we will consider to be perfectly accurate), a time-consuming and expensive test for SFI. When a patient is said to be SFI-positive, we mean that the time-consuming and expensive test has determined a patient to be SFI-positive.

Assume a simple random sample of 100 people is selected to be tested. Using the time-consuming and expensive test, it is determined that 32 of them have SFI, but the CHEAP test only found 30 of these. Similarly, CHEAP reported a “negative” result to only 61 of the 68 SFI-negative people.

1. (2 points) The two-way table below has the appropriate counts from these 100 patients. Also fill in the appropriate initials for TP, TN, FN, and FP from the definitions above.

2. (1 points) Enter this data into SPSS to get a table of the counts. (It should look very much like the table above.)

3. (1 point) Have SPSS produce a table with the joint and marginal percentages.

4. (2 points) What is the percentage of all patients who received a CHEAP positive test result? Is this a joint, marginal, or conditional percentage?

5. (2 points) Out of all the patients given the CHEAP test, what is the percentage of false negatives? Is this a joint, marginal, or conditional percentage?

6. Your roommate is unsure of his SFI status and gets the CHEAP test. His parents are devastated to hear that he has a positive test result. Knowing some statistics, you want to present him with a good reason why they should not be so upset yet. For patients who receive a positive CHEAP test result, what is the probability that they actually are SFI-positive? Is this a joint, marginal, or conditional percentage? Also write a note to your roommate explaining why your calculation should reassure him, and tell him what he should do now. Use at least one of the terms defined in the definitions box on page 1.

7. (2 points) What is the sensitivity rate? Is this a joint, marginal, or conditional percentage?

8. (4 points) Is there an association between the CHEAP test results and actual SFI status? (For a really poor test, the answer would be “no.” Think about it.) State your hypotheses, use SPSS to perform a chi-squared test, and give your test statistic and P-value. (Use an =0.01.) State your conclusion in terms of the problem.

9. (2 points) Is it appropriate to do a chi-squared test here? Why or why not?

SPSS Instructions for Lab 9

1. Input data into SPSS Data Editor. “CHEAPtest” is the row variable (string), “ActualSFI” is the column variable (string), and “Count” is the cell count (numeric, 0 decimal places). You may change variable names and decimals by clicking on the button “Variable View” at the left bottom corner of SPSS Data Editor.

Your data set should look like:

2. Weight Cases. Click on Data Weight Cases. Then click on “Weight cases by” and move “count” in to the “frequency variable” box. Then click on “OK”.

3. Create tables.

a. Create a count table.

Analyze Descriptive Statistics Crosstabs. Move “CHEAPtest” into the “Row(s)” box and “ActualSFI” into the “Column(s)” box. Then click on “OK”.

b. Create a percentage table and perform a Chi-square test.

· Analyze Descriptive Statistics Crosstabs.

· Move “CHEAPtest” into the “Row(s)” box and “ActualSFI” into the “Column(s)” box.

· Click on “Statistics” at the bottom of the Crosstabs dialogue box. Check “Chi-square” in the new dialogue box. Then “Continue”.

· Click on “Cells” at the bottom of the Crosstabs dialogue box. Check “Total” box under “Percentages” in the new dialogue box. Then “Continue” “OK”.

Note: The output tables for problem 1 might be too large to fit one page. You may right click on the table in the SPSS output window, choose “Copy objects” (not just “Copy”), and then paste it into Microsoft Word, so that you can adjust the size of the table.

1