Using the BRFSS dataset, conduct one-sample proportions to obtain the proportion and the 95% Confidence Intervals of that proportion. Refer to SPSS Lab #3 lecture for a step-by-step process. In general, you will do the following steps for each of these questions:

Public Relations by Geoffrey

Lab Assignment

Population Health Outcomes: Interpreting Results for Better Science

During this module, we have focused on the importance of understanding populations and the factors affecting that population – from cultural to health disparities – to improve health outcomes. We have also highlighted the importance of data in guiding your approach.  We have talked about the theoretical constructs and practiced doing calculations by hand.  Now, it is time to go deeper into the data and to practice analyzing actual datasets.

Instructions:

You will continue to use the BRFSS dataset to answer the questions in this assignment.

Prior to completing this lesson, be sure to review:

  • SPSS Lab #3 guided lectures on Canvas.
  • SPSS Survival Manual Ch. 8 (for the additional activity)

Recommend that you work through this assignment first, and then answer the questions on Canvas using the results you have found.

Please read instructions to each section carefully to understand what type of response you should be giving. This should illustrate that you completed the work yourself and using SPSS.

Questions – One-sample t-tests

Using the BRFSS dataset, conduct one-sample t-tests to obtain the Mean and the 95% Confidence Intervals of the Mean. Refer to SPSS Lab #3 lecture for a step-by-step process. In general, you will do the following steps for each of these questions:

  1. Analyze > Compare Means > One-Sample T Test.
  2. Select your variable(s) and move them into the Test Variable(s) window.
  3. Click Options and verify that you have the desired confidence interval percentage entered.
  4. Click Ok or Paste (and then highlight and run the syntax).
  5. Report the Mean (which is also shown in the Mean Difference column) and the Lower and Upper 95% CI values for the mean.

These 4 variables will be asked in the quiz for credit. Make sure to copy directly from the output and not to round.

Sleeptime Age Strength Height

 

Questions – One-sample proportions

Using the BRFSS dataset, conduct one-sample proportions to obtain the proportion and the 95% Confidence Intervals of that proportion. Refer to SPSS Lab #3 lecture for a step-by-step process. In general, you will do the following steps for each of these questions:

  1. Analyze > Compare Means > One-Sample Proportions.
  2. Select your variable(s) and move them into the Test Variable(s) window.
  3. In Define Success, select the value choice for the group you wish to know the proportion and the confidence interval. If you know that it is the first or last value, then you can select the corresponding options; otherwise type the numeric value in the Value box.
  4. Click Confidence Intervals and verify the desired confidence interval percentage entered in Coverage Level. Select ONLY “Clopper-Pearson (Exact)” for Interval Type.
  5. Click Tests and select “None” under Test Type.
  6. Click Ok or Paste (and then highlight and run the syntax).
  7. Report the Proportion and the Lower and Upper 95% CI values for the proportion. Make sure you have the right group in the first column you are reporting on.

These 4 variables will be asked in the quiz for credit; the format is “Variable name = value for the proportion.” Make sure to copy directly from the output and not to round.

Sex = Male Smoke100 = Yes FluShot = Yes HIVTest = Yes

 

Additional Activity (not graded)

In SPSS Assignment #5, you will be required to create new variables for your analysis using existing variables in your data. Ch. 8 in the SPSS Survival Manual will be useful.

After you create a new variable, you should run descriptive statistics/frequency tables, histograms, and boxplots to verify that it worked properly.

  1. Create a new variable using calculations of values from other variables
    Using the Height and Weight variables, create a new variable for Body Mass Index (BMI). Limit it to no more than 2 decimal places—either use the RND function or edit the variable length (when you run the descriptives only 2 decimal places should be seen).

 

 

 

  1. Create a categorical variable from a continuous variable
    1. Using your new BMI variable, create a categorical variable based on the following table:
Lower Bound Upper Bound Value Value Label
Lowest value < 18.5 1 Underweight
>= 18.5 < 25 2 Normal
>= 25 < 30 3 Overweight
>= 30 Highest value 4 Obese

 

  1. A dichotomous age variable where it is split as < 65 and >= 65.

 

  1. Combine 2 categorical variables into 1 new categorical variable
    1. The current race variable in the BRFSS dataset has too many categories to be useful in analysis. A modified combination variable of race and ethnicity is shown in the table below. Create this new summary variable using the Race and HispLat variables. (Hint: it may be useful to assess Hispanic or Latino ethnicity first and then look at the Race variable second.)
Hisplat Race New Variable Value Value Label
2 10 1 White, non-Hispanic
2 20 2 Black, non-Hispanic
1 Any value 3 Hispanic/Latino
2 30, 40, 41, 42, 43, 44, 45, 46, 47, 50, 51, 52, 53, 54, 60, 70 4 Other, non-Hispanic

 

 

  1. There are a few smoking variables and the skip-patterns of the questions means that some of the questions have missing values due to the answers on a prior question. Smoke100 asked everyone if they had ever smoked a certain amount (yes or no). Smokeday asked how often a person smoked, but it was skipped if the person had answered “no” to the Smoke100 question. You need to create a 3-category smoking variable as shown below that will combine these two variables.
Smoke100 Smokeday New Variable Value Value Label
2 Missing 1 Never Smoker
1 3 2 Former Smoker
1 1 or 2 3 Current Smoker