----- Original Message -----
From: "Copeland, Laurel" <[email protected]>
To: <[email protected]>
Sent: Monday, June 16, 2003 7:59 AM
Subject: st: test for trend from table data?
> A pharmacist asked me if I could do a test for trend over time from this
> table:
>
> Year Pharm Clin
> 1997 400 44
> 1998 221 54
> 1999 187 69
> 2000 123 63
> 2001 74 42
> 2002 145 149
>
> The table represents the number of reports of adverse events made by
> pharmacists (Pharm) and by clinicians (Clin), respectively. She observes
> that the number of reports declined until 2002 (there was an intervention
at
> the beginning of 2002), and that the proportion of reports made by
> pharmacists seems to dwindle. She wants a test statistic for either of
> these observations.
>
> Do I need to -expand- by pharm and clin? Or is there an immediate test
that
> might work? I am having trouble seeing how to approach this.
>
> All ideas are welcome!
>
> Thank you,
> Laurel
>
-findit trend- brings several options. In particular you may be interest in
the FAQ "Does Stata provide a test for trend?"
(http://www.stata.com/support/faqs/stat/trend.html) and Patrick
Royston's -ptrend- for trend analysis of proportions.
-ptrend- would give you this:
. ptrend phar clin year
pharm clin _prop year
|-----------------------------|
1. 400 44 0.901 1997
2. 221 54 0.804 1998
3. 187 69 0.730 1999
4. 123 63 0.661 2000
5. 74 42 0.638 2001
6. 145 149 0.493 2002
Trend analysis for proportions
Regression of p = pharm/(pharm+clin) on year:
Slope = -.07784, std. error = .00606, Z = 12.837
Overall chi2(5) = 167.206, pr>chi2 = 0.0000
Chi2(1) for trend = 164.789, pr>chi2 = 0.0000
Chi2(4) for departure = 2.416, pr>chi2 = 0.6597
Hope this helps,
Scott
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
