Calculate the mean of a frequency distribution on an HP 48SX?
September 27, 2006 8:12 PM   Subscribe

How do I calculate the mean of a frequency distribution on an HP 48SX calculator?

My wife is in a statistics class, and we're trying to figure out how to do problems with an old HP 48SX calculator that neither of us really know how to use well. Her problem is to calculate the mean of the following frequency distribution of the class and x is the class midpoint:

age ranges: 21-30, 31-40, 41-50, 51-60, 61-70, 71-80
frequency (f): 28, 30, 12, 2, 2, 2
midpoints (x): 25.5, 35.5, 45.5, 55.5, 65.5, 75.5

I don't know jack about stats, and she's just learning, but the answer figured by hand and verified in class is 35.76. How do we do this in the calculator? We figured out how to enter the data, which looks like this:

(sigma)DAT(1)=[25.5 35.5 45.5 55.5, 65.5 75.5]
(sigma)DAT(2)=[28 30 12 2 2 2]

Pressing the MEAN button gives us [26.75 32.75...]. How do we get the answer 35.76 instead of the mean of each pair of values? The instructions given in class were for a TI-84, not an HP. Help!
posted by DakotaPaul to Education (6 answers total)
 
To calculate the mean you have to apply the formula:
[Sum of (Midpoint * Frequency)] / [Sum of Frequency]

In this case this gives:
Sum of (Midpoint * Frequency) = 28*25.5 + 30*35.5 + 12*45.5 + 2*55.5 + 2*65.5 + 2*75.5 = 2718

Sum of Frequency = 28 + 30 + 12 + 2 + 2 + 2 = 76

2718 / 76 = 35.76

I would recommend that she learn the formula instead of the function buttons on the calc, it's better for actual learning of statistical math plus she won;t be impaired when she switches calculators.
posted by Vindaloo at 9:04 PM on September 27, 2006


Response by poster: She has the formula -- that's how she got the anser to begin with -- but this was a small problem to work with. Some others in her book have many, many more midpoints and frequencies and it would be nice to be able to punch them all in and hit MEAN to verify the answer.
posted by DakotaPaul at 8:21 AM on September 28, 2006


Looking through the HP48SX manual I didn't find anything about frequency distributions.

You might check out hpcalc.org's collection of software for the HP48 series. There are several statistics packages that might have what you want although I didn't see anything that leapt out at me.

Alternatively, you could write a program, since the formula doesn't look too complicated. That same site should have some good information on how to write programs for the calculator.

Be warned though: the main language for writing user programs on the HP48 series is a variant of Lisp that you might find a bit hard to wrap your head around if you don't have any experience with that kind of programming.
posted by jedicus at 8:44 AM on September 28, 2006


Response by poster: Thanks, jedicus, hpcalc.org looks very useful. Unfortunately, we don't have the cable to connect the calculator to a computer and we have a Mac anyway. If all else fails, we'll try to write the program. Is Lisp anything like C? ;-)
posted by DakotaPaul at 9:49 AM on September 28, 2006


You might find the RPL tutorials here and here useful. RPL is 'Reverse Polish Lisp,' the Lisp dialect used by the HP48 series.

It also appears that with a bit of work it is possible to connect a modern Mac to an HP48. You'd need a USB-serial converter and the old serial connectivity kit as well as a terminal emulator capable of doing KERMIT file transfers.

The cables are still being sold (e.g., here), though they're a bit expensive at $17.99 + shipping (and that's a clearance price). Given that you probably need this stuff working sooner rather than later, you might have a go at writing a program rather than trying the cable route, especially since some people on newsgroups have reported trouble getting such a setup to work.
posted by jedicus at 10:21 AM on September 28, 2006


Response by poster: Wow, thanks for the info, jedicus!
posted by DakotaPaul at 11:31 AM on September 28, 2006


« Older What email scam is this?   |   Has anyone outdone Orson? Newer »
This thread is closed to new comments.