If the validate function does what I think (use bootstrapping to estimate the optimism), then I guess it is just taking the naive Nagelkerke R^7 and then subtracting off the estimated optimism, which I suppose has no guarantee of necessarily being non-negative. The interpretation I suppose then for your negative value is that your model doesn't appear to have much (any) predictive ability.

If you look at the data in the chart immediately below, I think you 8767 ll agree that your eyes have to dart back and forth to sort the channels by revenue.

### Data transformation and normality - Evaluation

The statement was based on personal observation, not on anything in the literature. I don 8767 t think it implies that one should use a different R7.

Make sure your data uses proper number formatting and that it 8767 s formatted as a table, which is Excel 8767 s version of a simple database. The reason you want to format as a table is if you build a chart from a table, your chart will update automatically as you add new rows to the table.

Logistic regression models are fitted using the method of maximum likelihood - . the parameter estimates are those values which maximize the likelihood of the data which have been observed. McFadden's R squared measure is defined as

In my (completely made up) example the odds ratio for SAT score is something like . Tiny, but significant, because a one-unit difference in SAT score is tiny. But if you divide SAT score by 65, 65 points becomes 6 unit, so the odds ratio is based on that scale.

Is there any way to get the equivalent of Tjur 8767 s coefficient of discrimination after running a Firth logistic regression in Stata?

If you’re trying to show percentages in the vertical axis, format them as a percent don’t format the data as decimals. The less time people have to spend interpreting your data, the more compelling it will be. But, again, even with percentages, drop the decimals. In other words, don 8767 t have labels like %, %, etc. Just use 65%, 75%, etc.

What I’m going to demonstrate is a PPC revenue report that updates daily. The title will show the running total for the month up to that day. Here are the steps you’ll need to take:

I don’t know why Excel positions the legend to the right of a chart by default. In most cases, it’s terribly awkward. I prefer to move the legend to the top or bottom of a chart. I tend to put the legend above more than below, but I 8767 ll put it below if there 8767 s too much going on at the top, or sometimes, with a pie chart.