)/cmu-logo.jpg "Carnegie Mellon"){kind=logo}
Tutorial 3: Searching Using One Variable
A search using one input criteria is the most straightforward to do and the easiest to interpret. Starting on the “Death Comparison” webpage, you select one of seven options in “Step 1: Primary Comparison” (See Figure 3-1). For example, to compare the risk of dying for males and females, select “Gender.” To compare the risks for different age groups, select “Age.”
)/images/figure31.jpg "Figure 3-1")
Figure 3-1: Choosing primary category
A basic search of immediate interest might be to find out your own risk of dying in the next year and from what cause. DeathRiskRankings.org cannot give you your exact profile, but it can provide the average risk for someone of your age, gender, race (in the US), and region. The search procedure is straightforward.
- Step 1: Primary Comparison: Select Causes of Death (Figure 3-1)
- Step 2: Secondary Comparison: Select No Secondary Comparison (more on the other options in Step 2 in the next tutorial).
- Step 3: Search Criteria: Select specific demographics about the person of interest (yourself). (Figure 3-2).
- Age: 21 (Note: individual ages are found after the age categories on the pull-down menu)
- Gender: Male
- US or Europe: United States
- Region: Pennsylvania
- Race: White (Note: race information is only available for the US comparisons)
- Cause of Death: Not selectable because it’s selected in Step 1
- Simple Language: Use the default, that is the less technical medical terms
- Forecast Period: Dying within 1 year
- Step 4: Submit: Click on the Submit button to complete the search (Figure 3-3)
)/images/figure32.jpg "Figure 3-2")
Figure 3-2: Search details
)/images/figure33.jpg "Figure 3-3")
Figure 3-3: Completing the search
The search results will appear in the same window below the input options. A summary of the search will appear to make sure that the output displayed matches the search that you wanted and did not make a mistake (Figure 3-4). The Search Parameters table boxed in red, shows the specifics of your search. Using the buttons in the Display Options box, you can change the metric with which your results are displayed. These options will be discussed in Tutorial 6. For this first example, we use the default settings.
)/images/figure34.jpg "Figure 3-4")
Figure 3-4: Output information
Figure 3-5 shows the output table for this search. The default output metric is the MicroMort, or one-in-a-million chance of dying. For more details on MicroMorts see Tutorial 2. A simple guiding principle is that MicroMorts are bad and that having a lot is not good.
The output table shows that for a 21-year old, white, male living in Pennsylvania, the chance of dying next year is 1,320 MicroMorts. So, if there were 1 million white 21-year olds in Pennsylvania, 1,320 would die next year. This is approximately a 1-in-750 chance of dying (that is 1,000,000/1,320 = 757). In percentage terms, there is a 0.13% chance of dying next year (1,320/1,000,000).
The table also shows details as to the causes of this chance of dying. The default display lists up to 20 summary categories of causes in alphabetical order and their associated MicroMorts. The sum of all the causes will equal the total. To sort based on MicroMorts, simply click on the MicroMort header in the table. Multiple clicks will alternate the sort between increasing or decreasing in value. Clicking on the Cause of Death header will sort the table alphabetically by the causes. By visually scanning down the list, it is easy to pick out the top causes: Accidents (with 757 MicroMorts), Suicides (with 213 MicroMorts), and Homicide (with 81 MicroMorts).
)/images/figure35.jpg "Figure 3-5")
Figure 3-5: MicroMort output display
It is interesting to note that these three top causes (totaling 1,051 MicroMorts or 80% of the total of 1,320) have nothing to do with disease. Cancers are the fourth leading cause with 57 MicroMorts (comprising approximately 4% of total mortality risk); however, 57 MicroMorts translates to about a 1 in 17,500 chance of dying in the coming year. This is just as likely as flipping 16 fair coins and having all of them come up heads.
This single variable search allows the user to explore death risk by one dimension. However, in many cases, you want to know how risks compare between different groups (for example, males versus females, or US versus Europeans). To do this, you need to run a two-variable search, which is described in Tutorial 4.