Based on self-reporting of symptoms and conditions.
Only items with 16 or more reports in the current filter are shown. 16 is the minimal level to detect significance.
What are we doing? Imagine taking all of the salaries in a company and grouping them into the 4 EQUAL-SIZED groups - creating brackets (0-$18000, $18000-$63000, $63000- $75000, over $75000).
We now do a count of women in each of these brackets. We now look for counts in each bracket that are above or below expectation. If we see 75% in $18000-$63000 (instead of the expected 25%), we would conclude that there may be inequality in wages for women. In some cases, the average women wage may be the same as men (because a few of the highest paid are women, bringing up the average)
In this case, it may be that women are not hired for maintenance, grounds (lowest bracket) or managers -- just as clerical. This is how we detect abnormalities with bacteria. Are there A LOT more or less than expected in one or more of the brackets?
Only statistically significant patterns are shown below. The number in each Quartile are expected to be the same. Observed for these conditions is shown.
|Bacteria||People||Expected in each group||Bottom 25%||Low 25%||High 25%||Top 25%||P Value||Eureka!|
Warning: By random chance, 5% (PValue 0.05) are likely to be reported. So if 1000 bacteria are checked, 50 will likely be reported as "0.05 significant" (5% of 1000). The Eureka column shows those that are likely to be meaningful.
The Eureka PValues is about 0.00143472022955524 because we have 348 bacteria examined.