An official website of the Department of Health & Human Services
Quick Guide to MEPSnet/HC
The data source selection tab allows you to select a data year to use in your analysis. Each December the most recent released full-year consolidated data file is added to the set of data years available for analysis with MEPSnet HC.
Full-year person-level files are available containing detailed data on survey administration, demographics, employment, health status, health insurance, income, and health care utilization and expenditures.
Beside each file is the image, , which is linked to page where you can obtain documentation and a codebook for this data file.
The variable selection tab allows you to select a variables to use in your analysis. These variables may be used for the following purposes on other screens.
On this screen, you do not need to indicate how these variables are to be used. This is done on other screens. You also do not need to use every variable that you select in your analysis. If you are unsure if you will want to use a variable, there is no harm in selecting it. If you move on to another screen and decide that you would like to use variables that you have not selected, you will be able to come back to this screen to select more variables. If you would like to perform a weighted analysis or calculate standard errors, there is no need to select weight or variance estimation variables. These variables will automatically be used.
The variables are organized into groups. To select variables in a group click the link for the group. When you click a link for a group, instead of seeing variables in the group, a list of subgroups may be listed below the group. To see the variables in a subgroup, click the link for the subgroup.
To select a variable, check the check box that preceedes it. If you would like to use variables from a different group also, select that group and select more variables. The variables that you had already selected will remain selected, and you will see these variables below the list of variables in the new variable group. To deselect a variable, uncheck the check box that preceeds it. Any number of variables may be selected.
You select the hyperlinked variable name to obtain a frequency of the variable. If the variable is continuous, you will also see the mean, median, and quartiles for the variable. There may be a few second pause before you see the statistics.
The variable recoding tab allows you to recode a variables that you have selected for use in your analysis. The variable recoding will be used to group values of a variable into categories that will be used when the variable is used as a classification variable in your analysis.
To recode a variable, select the link for the variable in the variable pick list. When you do, you will see a frequency for that variable using the current coding for that variable. If you have not changed the recoding for the variable, a frequecncy for the default recoding will be shown.
Following the frequency, you will see several controls to change the recoding for the variable. On each line, you can provide a range of values for the variables and a label for that range. For each range, you can specify a start point and end point, and specify whether the start and end points should be included in the range.
You can remove a range by removing the label (making it blank). If there are not enough rows for all of the ranges that you would like to use, select Add Row.
Once you have made the changes that you want, select Update. Until you do this, your changes to the recoding of a variable will not be made--and if you proceed to another tab, your changes will be lost. When you select Update, your changes to the recoding are made, and you will see a new frequency of the variable. Any errors in the ranges (ranges that overlap, or are invalid) will be reported. You will also be notified if there are any values of the variable that you did not include in any range.
The ranges that you choose will be used the variable is selected as a classification variable. If more than one range has the same label, the two ranges will be grouped together.
The Record Selection tab can be used to select the records to be used in the analysis. By default, all records in the file that you chose will be used. You can use this screen to build an expression, which must be true for a record to be included in the analysis.
At the top of the screen, under Current Record Selection, you will see a description of the currently selected records. By default, all records are selected. When you make a change to the selected records, this description will be updated to reflect the new set of records that you have selected. This description will include the expression that you have given select the records, a count of the records selected, and the weighted population selected.
Below this, under New Record Selection are some controls that can be used make an expression to change the selected records. The expression is built by making comparisons between variables and values that the variable may contain. Each row may contain up to two comparisons. If the left and right sided of a comparison are left blank, the comparison is omitted from the expression. When both comparisons are specified for a row, you may specify that the row should be true if the first AND the second comparison are true (both are true), or you may specify that the row should be true if the first OR the second comparison is true (either are true). If more than one row is used, the truth of each row is determined, then the truth of the expression as a whole is determined.
Each expression consists of three controls. On the left side is a dropdown box that may be used to select a variable, in the center is a dropdown box that may be used to select a relational operator, and on the right side of the expression is a box in which you can type a value (or values) of the variable. The relational operators that may be selected are as follows.
If you are unsure what the values of some variable are, you can find out by selecting (clicking) a hyperlinked variable name to obtain more information about the values of a variable.
Sometimes a single comparion is not adequate to identify the records that you want to use in the analysis. You can use up to six comparisons-two to a row and up to three rows. When you use more than one comparison, you need to indicate whether you want to include records where the comparisons are all true or where any comparisons are true. When you use both comparisons on a row, you may specify that the row should be true if the first AND the second comparison are true (both are true), or you may specify that the row should be true if the first OR the second comparison is true (either are true). If more than one row is used, the truth of each row is determined, then the truth of the expression as a whole is determined. Below the third row is a box allowing you to indicate if a record should be included in the analysis if the row 1, 2, AND 3 are true (all three are true), or if row 1, 2, OR 3 are true (any of the three are true).
Once you have made the changes that you want, select Update. Until you do this, your changes to the record selection will not be made--and if you proceed to another tab, your changes will be lost. When you select Update, your changes to the record selection are made, and you will see a new description of the record selection under Current Record Selection. Any errors in the expression that you made will be reported. This can include a missing variable or value in one of the six comparisons, or a non-numeric value in a comparison.
The Descriptive Statistics tab can be used to generate the descriptive statistics. You may use this screen to generate frequencies (and percentages) of records by one or more of the variables that you have selected. You can also generate the other statistics such as the mean, median, or quartiles of a continuous variable. The statistics will be weighted, and standard errors which take into account the sample design will be calculated and displayed along with the estimates.
In order to generate statistics, you need to choose a type of statistic by selecting (clicking) the link for that type of statistic. Once you do this, you will see a set of controls that can be used to generate statistics of the type that you selected. Fill out these controls to describe the statistics that you would like to see, and select (click) Show Statistics. (This will take anywhere from 5 seconds to a minute or more depending on the statistics requested and the number of requests that are being made at that time.) Once the statistics have been displayed, you can return to the Descriptive Statistics screen by selecting (clicking) More Statistics.
Types of Statistics
The following types of statisitics can be selected.
One Way Frequencies
The One Way Frequencies option will allow you to generate population counts for each of the possible values of a variable.
Up to six variables can be selected in the drop-down boxes provided. A one-way frequency will be generated for each of the variables that you select. Each frequency will show the population count for each of the possible values of the variable. The values of a variable may have been recoded into categories. In this case, population counts will be generated for each of these categories. These categories can be changed, if desired, using the Variable Recoding tab.
To generate the statistics, select (click) Show Statistics. (This will take anywhere from 5 seconds to a minute or more depending on the statistics requested and the number of requests that are being made at that time.) Once the statistics have been displayed, you can return to the Descriptive Statistics screen by selecting (clicking) More Statistics.
n-Way Frequencies
The n-Way Frequencies option will allow you to generate a frequency crossing each of the variables that you select. The frequency will show the population count for each of the possible combinations of the values of the variables.
Up to three variables can be selected in the drop-down boxes provided. A frequency will be generated crossing each of the variables that you select. The frequency will show the population count for each of the possible combinations of the values of the variables. The values of a variable may have been recoded into categories. In this case, population counts will be generated for each of the possible values of these categories. These categories can be changed, if desired, using the Variable Recoding tab.
Minimums, Maximums, Sums, Means, and Medians
The Minimums, Maximums, Sums, Means, and Medians option will allow you to generate minimums, maximums, sums, means, and/or medians of a variable. These statistics can be obtained separately for each value (or category) of one or more other variables.
Select a variable for which you would like to generate statistics. Then, check the minimum, maximum, sum, mean, and/or median checkboxes to select the statistics that you would like to see. (Any number may be selected.)
You can generate these statistics for the entire population or for each of the possible combinations of the values of up to three variables. These variables can be selected in the drop-down boxes near the bottom of the screen. The values of a variable may have been recoded into categories. In this case, statistics will be generated for each of the possible values of these categories. These categories can be changed, if desired, using the Variable Recoding tab. To generate statistic for the entire population, do not select any of these variables.
Ratio Estimates
The Ratio Estimates option will allow you to generate ratios of one variable divided by another variable. Ratios can be obtained separately for each value (or category) of one or more other variables.
In the dropdown boxes near the top of the screen, a variable which you would like to use as the numerator of the ratio and the variable that you would like to use as the denominator of the ratio.
Advanced
The Advanced option provides a compact interface that can generate any of the statistics above.
Frequencies can be generated by first selecting one or more classification variables. If you would like a single frequency for each classification variable, leave the Cross the classification variables box unchecked, or check this box for cross-frequencies. Make sure that --None-- is selected in the Analysis Variable drop down box. If you would like the analysis to be weighted, select the Weight by Person-Level Weight box. If you would like standard errors to be calculated, select the Calculate Standard Errors box. To generate the statistics, select (click) Show Statistics. (This will take anywhere from 5 seconds to a minute or more depending on the statistics requested and the number of requests that are being made at that time.)
The minimum, maximum, sum, mean, median, quartiles, and deciles of a variable may be calculated as well as the ratio the variable to another variable. To do this select the variable that you would like to use to calculate these statistics in the Analysis Variable drop down box. Select the statistics that you would like to see by checking the box beside each statistic. For the ratio statistic, you must also select a variable to use as the denominator of the ratio. Classification variables can be selected to provide a separate set of statistics for each value of a classification variable. If more than one classification variable is selected, you may choose to obtain statistics for each value classification variable separately, or to cross the classification variables. If you would like the analysis to be weighted, select the Weight by Person-Level Weight box. If you would like standard errors to be calculated, select the Calculate Standard Errors box. To generate the statistics, select (click) Show Statistics. (This will take anywhere from 5 seconds to a minute or more depending on the statistics requested and the number of requests that are being made at that time.)
Once the statistics have been displayed, you can return to the Descriptive Statistics screen by selecting (clicking) More Statistics.