Bugs fixed in 1.1 Aug 11, 2000
- Dan Talberg Reported: I was looking at your java applet 'histogram explorer' in the resource center of Rice virtual lab in statistics. The applet opens with all of the bars of the histogram at 5.
When the bar at a score of 1 is moved up, the mean moves down, as expected. But when the '1' bar is lifted to a frequency of 10, the mean jumps up to 6. It's as if the
applet sees the 10 as a value of 0. Just thought I'd let you know to avoid confusing students.
- Turns out that my number formatting routine was doing bad things. Any number that had a zero in the first decimal place (5.09) got the zero dropped (5.9). So this bug was pretty bad. Thankfully Java has these routines built in now. I've switched to it and it seems fine.
- You can now download the source, class and HTML files
Bugs...Bugs...Bugs in version 1.01
Most of the bugs are related to how things get displayed. I'm please at this point that it works at all
- On Netscape 3.X PC -- The entry fields are always showing.
- resizing the screen will sometimes make this go away, but don't maximize the screen as it will looks silly. Works ok on Internet Explorer 3.X
- Sometimes the Check Guesses and Clear Guesses button get half way off the screen.
- I clicked the close button on the window and now the button says ...Running... instead of begin
- The ...Running... button just means that the histogram window is already open. Sometimes the window gets hidden or killed and the button doesn't switch back. Force this page to reload by holding down the shift key on PCs and UNIX boxes and the option key on macs and then click the reload button. This does a "super" reload and resets the applet.
- Ron Ketter Reports "One note -- when I redraw the histogram to create bimodal distributions
with gaps in the center [e.g., f(2)=3, f(3)=5, f(4)=3, f(8)=3, f(9)=5,
f(10)=3], it shows the median at 4.5 instead of 6. I've tried this with
several bimodal distributions, and the median appears at the right edge
of the left grouping instead of the center."
I had calculated the median and IQ range wrong. The case where
number of scores is even was not averaging the two middle scores.
- Anthony Villegas--Everything seems to be correct except for the left-skewed and right-skewed distribution
which are vice-versa.
Skew means tail and left-skewed means the tail of the distribution
points to the left (negative side), so right-skewed means the tail
points to the right (positive side).
- Switched em.. eds
If you like you can send comments and bug reports to Eric Scheide.