BookmarkSubscribeRSS Feed
Tamino
Obsidian | Level 7

Hello

 

I am doing a interrupted time series analysis / autogregression and I am not sure about the interpretation of the output.

My dataset / variables looks like this:

 

period ddd     denominator prop_ddd swissmedic swissmedic_per
1      1125320 1115442     10088.56     0             0
2      1227815 1115442     11007.43     0             0
3      1315611 1115442     11794.53     0             0
4      1287360 1115442     11541.25     0             0
...
60     1243608 1077136     11545.51     1             .
61     1111666 1084254     10252.82     1             1
62      929344 1130519      8220.51     1             2
63     1031575 1130519      9124.79     1             3
...
85     1160896 1241240      9352.71     1             25

 

and my comand is following:

 

proc autoreg data=bpmedi.autoreg_ddd_hct;
model prop_ddd = period swissmedic swissmed_per / method=ml nlag=1 dwprob
covb;
output out=autoreg_out predicted=pred pm=trend lcl=pred_l ucl=pred_u residual=resid
        predictedm=predm lclm=predm_l uclm=predm_u residualm=residm;
run;

 

 And i get the output attached as screenshots.

 

I want to know the trend before the intervention (in period 60), the immediate effect after the invention and the trend after the intervention. My understanding is, that I need the table with "Parameter Estimates", but why is my intercept not the value of the pro_ddd in period 1 (11627 vs 10088.56)?

 

Thank you for your help!Autoreg1.jpgAutoreg2.jpg

sas-innovate-white.png

Missed SAS Innovate in Orlando?

Catch the best of SAS Innovate 2025 — anytime, anywhere. Stream powerful keynotes, real-world demos, and game-changing insights from the world’s leading data and AI minds.

 

Register now

Discussion stats
  • 0 replies
  • 722 views
  • 0 likes
  • 1 in conversation