Topic of study is "An empirical study on factor affecting operational efficiency of commercial banks". So in our study we have taken 14 Independent factors and 8 dependent factors. Factors are as follows
Independent factors are: 1 Total assets, 2 Advances, 3 Deposits, 4 Borrowings, 5 Fixed Assets, 6 Operating Cost, 7 Staff cost, 8 NPA, 9 Cash Deposit Ratio, 10 Investment deposit Ratio, 11 Interest Expenses, 12 Business /Employee, 13 Liquidity, 14 Credit Deposit Ratio
Dependent Factors are: 1 ROA, 2 ROE, 3 Return on Net worth, 4 Net Profit, 5 Interest earned 6 Interest Expended, 7 Capital adequacy 8 Profit per employee
Problems identified:
Usually we study the relationship between 1 dependent and multiple independent factors.
Can we have any method of study where we can use multiple dependent factors in collective manner rather than going one dependent factor at a time. To check the relation between the various dependent factors also in a common equation.
I would go for artificial neural networks. You can get information about the individual contribution of each predictor and also how they correlate amongst them (classification, as in factor analysis) to influence the outcomes.
Dear Anoop,
You can look at the regression analyses our paper on Depression, anxiety, hostility and hysterectomy. If you have any questions consult Lertola here on RG.
Actually there is no problem with many IV:s and DV:s. However check for multicollinearity in your IV:s and omit those below ,3 and above .7. SPSS checks the Tolerance and VIF ( Variance inflation factor) values for you. If Tolerance is less than .10 then multiple correlations with other variables is high and if VIF is greater than 10 you have to omit some variables.
I think that you have to perform more than one regression analysis based on different interactions between 14 IV: and 9 DV:s. Your research will benefit from it.
In SPSS you could do ANALYZE - GENERAL LINEAR MODEL - MULTIVARIATE then you move your dependent variables into the DEPENDENT VARIABLES box, your categorical independent variables into the FIXED FACTORS box and your continuous independent variables into the COVARIATES box. Under OPTIONS you have to specify that you want the regression coefficients.
I will run a MANOVA (or MANCOVA if there are confunding variables involved) instead. But if you want to continue with a regression model, then make sure that all the assumptions are satisfied.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Applied Statistics