topic Re: Errors in regression model with interaction term in SAS Studio

Errors in regression model with interaction term

sastuck — Mon, 02 Apr 2018 15:03:49 GMT

Hello,

@Kurt_Bremser and @art297 helped me with some code to define a dummy variable which was 1 if a company's annual return had increased from the previous year and 0 otherwise.

I want to interact this term with annual_return in my model so that I can test the hypothesis that executives will be rewarded when the company does well but to be insulated from downturns in performance (if the compensation contract is constructed to encourage risk-taking behavior).

Running what I thought to be the appropriate model, however, I ran into two error messages, which I don't understand. Here is the log:

1 OPTIONS NONOTES NOSTIMER NOSOURCE NOSYNTAXCHECK;

NOTE: ODS statements in the SAS Studio environment may disable some output features.

72 ***OLS REGRESSION***

73 title "";

74 ods graphics on;

75 proc reg data=paper.ceo_firm plots(maxpoints=none);

76 model salary = annual_return annual_return*increase/clb acov;

200

NOTE: The previous statement has been deleted.

ERROR 22-322: Syntax error, expecting one of the following: a name, ;, -, /, :, _ALL_, _CHARACTER_, _CHAR_, _NUMERIC_, {.

ERROR 200-322: The symbol is not recognized and will be ignored.

77 run;

WARNING: No variables specified for an SSCP matrix. Execution terminating.

NOTE: PROCEDURE REG used (Total process time):

real time 0.00 seconds

user cpu time 0.00 seconds

system cpu time 0.00 seconds

memory 3410.18k

OS Memory 34752.00k

Timestamp 04/02/2018 02:56:14 PM

Step Count 77 Switch Count 0

Page Faults 0

Page Reclaims 473

Page Swaps 0

Voluntary Context Switches 9

Involuntary Context Switches 0

Block Input Operations 0

Block Output Operations 48

78 ods graphics off;

80 OPTIONS NONOTES NOSTIMER NOSOURCE NOSYNTAXCHECK;

any help is appreciated!

Re: Errors in regression model with interaction term

PaigeMiller — Mon, 02 Apr 2018 15:28:51 GMT

PROC REG does not allow interactions via the X1*X2 syntax.

Use PROC GLM instead. (And in PROC GLM, you don't need to create a column of dummy variables, you can use the CLASS statement instead).

Re: Errors in regression model with interaction term

sastuck — Mon, 02 Apr 2018 15:50:11 GMT

is there more to do than just substituting reg with glm?

here's my new log.

1 OPTIONS NONOTES NOSTIMER NOSOURCE NOSYNTAXCHECK;

NOTE: ODS statements in the SAS Studio environment may disable some output features.

72 ***OLS REGRESSION***

73 title "";

74 ods graphics on;

75 proc glm data=paper.ceo_firm plots(maxpoints=none);

76 model salary = annual_return annual_return*increase/clb acov;

___

202

ERROR: Variable increase not found.

NOTE: The previous statement has been deleted.

ERROR 22-322: Syntax error, expecting one of the following: ;, ALIASING, ALPHA, CLI, CLM, CLPARM, COVBYCLASS, E, E1, E2, E3, E4,

EST, I, INTERCEPT, INVERSE, NOINT, NOUNI, P, PREDICTED, SINGULAR, SOLUTION, SS1, SS2, SS3, SS4, TOLERANCE, X, XPX,

ZETA.

ERROR 202-322: The option or parameter is not recognized and will be ignored.

77 run;

78 ods graphics off;

80 OPTIONS NONOTES NOSTIMER NOSOURCE NOSYNTAXCHECK;

and how does the class statement work?

Thanks!

Re: Errors in regression model with interaction term

PaigeMiller — Mon, 02 Apr 2018 15:54:16 GMT

The error message says that INCREASE is not a variable in data set paper.ceo_firm.

So you can't use INCREASE until you create it in paper.ceo_firm

Re: Errors in regression model with interaction term

sastuck — Mon, 02 Apr 2018 15:56:32 GMT

Ah, I didn't realize I had to re-run the datastep after logging back into SAS. The other error messages are still present, however.

Re: Errors in regression model with interaction term

PaigeMiller — Mon, 02 Apr 2018 16:49:05 GMT

Show us the SASLOG so we can see these other error messages after you fixed the first error.

Re: Errors in regression model with interaction term

sastuck — Mon, 02 Apr 2018 17:00:23 GMT

Here you are:

1 OPTIONS NONOTES NOSTIMER NOSOURCE NOSYNTAXCHECK;

NOTE: ODS statements in the SAS Studio environment may disable some output features.

72 ***OLS REGRESSION***

73 title "";

74 ods graphics on;

75 proc glm data=paper.ceo_firm plots(maxpoints=none);

76 model salary = annual_return annual_return*increase/clb acov;

___

202

NOTE: The previous statement has been deleted.

ERROR 22-322: Syntax error, expecting one of the following: ;, ALIASING, ALPHA, CLI, CLM, CLPARM, COVBYCLASS, E, E1, E2, E3, E4,

EST, I, INTERCEPT, INVERSE, NOINT, NOUNI, P, PREDICTED, SINGULAR, SOLUTION, SS1, SS2, SS3, SS4, TOLERANCE, X, XPX,

ZETA.

ERROR 202-322: The option or parameter is not recognized and will be ignored.

77 run;

78 ods graphics off;

80 OPTIONS NONOTES NOSTIMER NOSOURCE NOSYNTAXCHECK;

Re: Errors in regression model with interaction term

PaigeMiller — Mon, 02 Apr 2018 17:17:14 GMT

there is no CLB or ACOV option in PROC GLM.

From now on, please paste code and SASLOGs by first clicking on the "running Man" or "Insert SAS Code" icon. This will preserve spacing and make the code and SASLOGs much easier to read.

Re: Errors in regression model with interaction term

sastuck — Mon, 02 Apr 2018 17:20:24 GMT

Note taken. Thanks so much for the help!