Analysis Using SPSS  

To test the equality of treatment combinations and identification of best treatment combination requires analysis to be performed on treatment combinations. Therefore, 27 treatment combinations are recoded as:  

 Main Procedure is:  

Start →All Programs → SPSS for Windows → SPSS 15.0/ SPSS13.0/ SPSS10.0 (based on the version available on your machine) → Enter data in Data Editor → Analyze → GLM → Univariate → yield → [puts yield under Dependent list: ] → REP → [puts REP under Fixed Factor(s): ] → BLK→ [puts BLK under Fixed Factor(s): ]  V → [puts V under Fixed Factor(s):]  N → [puts N under Fixed Factor(s):] P → [puts P under Fixed Factor(s):] Continue → Model... [Opens Model dialogue box] → Custom → Build Term(s) → Main effects →   [puts V, N and P under Model:] → Interaction → [puts V*N V*P N*P and V*N*P under Model:] → Run All.

For performing analysis, input the data in the following format. 

{Here the Replication is termed as REP, three factors as V, N and P, treatment as TRT and block as BLK. It may, however, be noted that one can retain the same name or can code in any other fashion}.

Note: If one is interested in answering first two questions, then there is no need of recoding the treatment combinations and adding the variable TRT in the variable.

 Following are the brief description of the steps along with screen shots.

·         Open Data editor: Start → All Programs → SPSS for Windows → SPSS 15.0/ SPSS13.0/ SPSS10.0

·         Enter data in SPSS Data Editor. There are two views in SPSS Data Editor. In variable view, one can define the name of variables and variable types string or numeric and data view gives the spreadsheet in which data pertaining to variables may be entered in respective columns. In the present case, we enter data in numeric format.

·         Once the data entry is complete, Choose Analyze from the Menu Bar. Now select Analyze → General linear Model → Univariate.

·         This selection displays the following screen.

·         Select yield and send it to the Dependent Variable box; REP, BLK, V, N and P may be selected for Fixed Factor(s) box. After doing these the dialog box should be like this

·         For the Interactions select Model in the Univariate dialog box i.e. → Model... [Opens Model dialogue box] → Interaction→ V → N → [puts V*N under Model:]. Similarly V*P, N*P and V*N*P can also be put under model.

      This selection displays the following screen.

·         Click Continue to return to the Univariate dialog box.

·         All possible pair wise treatment comparisons can be performed using the Button Options on the dialogue box. A click on Button Options, gives the option for estimated marginal means and display means for. From the left hand box, take the effect treatment in the Display means for. Then check the box Compare main effects and then there are 3 options for confidence interval adjustment viz. LSD(none), Bonferrnoni and Sidak. Any one of these 3 options can be selected. Default option is LSD(None). 

       A screen shot for these options is

This, however, gives comparisons of single factor level means and provides only the means for 2 factor level combinations and 3-factor level combinations. For 2-factor and 3-factor level combinations pairwise comparisons or post hoc comparisons cannot be performed using SPSS.

·         Click  Continue to return to the Univariate dialog box and Click Paste to go to the Syntax editor mode and in the model define blk(rep)

• Click Run → All

For testing the significance of treatment combination, identification of best treatment combination use the following steps:

·         Choose Analyze from the Menu Bar. Now select Analyze → General linear Model → Univariate.

Select yield and send it to the Dependent Variable box; REP, BLK and TRT to the Fixed Factor(s) box.

·         Select Model in the Univariate dialog box i.e. → Model... [Opens Model dialogue box]. Put REP, BLK and TRT under the model for main effects.

·         Click Continue to return to the Univariate dialog box.

·         All possible pair wise treatment comparisons can be performed using the Button Options on the dialogue box. A click on Button Options, gives the option for estimated marginal means and display means for. From the left hand box, take the effect treatment in the Display means for. Then check the box Compare main effects and then there are 3 options for confidence interval adjustment viz. LSD(none), Bonferrnoni and Sidak. Any one of these 3 options can be selected. Default option is LSD(None). 

       A screen shot for these options is 

·         Click  Continue to return to the Univariate dialog box and Click Paste to go to the Syntax editor mode and in the model define blk(rep)

• Click Run → All

• To answer all the questions 1 to 3, the following syntax may be used after creating the data file.

UNIANOVA

  yield  BY rep blk v n p

  /METHOD = SSTYPE(3)

  /INTERCEPT = INCLUDE       

  /EMMEANS = TABLES(v) COMPARE ADJ(LSD)

  /EMMEANS = TABLES(n) COMPARE ADJ(LSD)

  /EMMEANS = TABLES(p) COMPARE ADJ(LSD)

  /EMMEANS = TABLES(n*v)

  /EMMEANS = TABLES(p*v)

  /EMMEANS = TABLES(n*p)

  /EMMEANS = TABLES(n*p*v)

  /CRITERIA = ALPHA(.05)

  /DESIGN = rep blk(rep) v n p n*v p*v n*p n*p*v .

UNIANOVA

  yield  BY rep blk trt

  /METHOD = SSTYPE(3)

  /INTERCEPT = INCLUDE

  /EMMEANS = TABLES(trt) COMPARE ADJ(LSD)

  /PRINT = DESCRIPTIVE

  /CRITERIA = ALPHA(.05)

  /DESIGN = rep blk(rep) trt .