I am currently studying conflict management strategies and teachers job effectiveness. The specific purpose of my study to find the relationship between
1. Arbitration strategy and teachers effectiveness
2. Dialogue and teachers effectiveness
3. Effective communication and .....
4. Smoothing strategy and ......
I will have used Pearson alone which am sure will do the work. I feel like finding the combine effect of arbitration, dialogue, effective communication, and smoothing on teachers effectiveness. Will it be nice to do this? Can I employ Pearson to run individual correlation based on the four areas, while I used multiple regression to combine all the 4 strategies on teachers effectiveness?
Dear Valentine
You can estimate the correlation coefficient ( Pearson correlation needs scale data and normal distributed data, while Spearman's rank correlation don't need these assumptions). Also, you can estimate the regression coefficient, but you must note that the correlation is the relationship between two independent variables, while the regression needs to use dependent variable and independent variable.
So, in this case you test different hypothesis in each estimation.
Good Luck
Hi Owan
1. Mathematically the multiple correlation coefficient could be calculated using multivariate normal distribution.See:
. Richard Johnson and Dearn Wichern (2007). Applied Multivariate Statistical Analysis. Chapter 4, pages: 149-208.
.Joseph Hair, Wuilliam Black, Barry Babin, Rolph Anderson and Ronald Tatham (2006).Multivariate Data Analysis. Chapter 4, pages: 237-243.
2. Using SAS. See, Willam Mendenhall and Terry Sincich (2003).Chapter 4, pages:182-185.
3. In SPSS, it can be calculated by R square using Multivariate ANOVA. See, Landau, S. and Everitt, B. S.(2004). A Handbook of Statistical analyses using SPSS. Chapter 4, pages:92-126.
Regards,
Zuhair
Hello Valentine,
I am interested to know how you measured your variables (e.g. Arbitration strategy, dialogue, effective communication etc.).
Yes, you can use both correlation and multiple regression to analyse your data. I suggest you estimate the correlation coeffficients and compare them with the results from your regression coefficients (for the same independent variables) and then comment on the changes observed. For example, the correlation coefficient between Arbitration Strategy and Teacher's Effectiveness (TE) may differ from the regression coefficient for Arbitration Strategy on TE.
Alternatively, if you want to analyse the direct and indirect effects of the independent variables on TE, you may carry out 'Path Analysis' (it will require drawing a path diagram). You can look it up, the methods are not too different from your usual correlation/regression.
Goodluck!
yes you can use both of simple correlation and regression (simple or multiple). The results depend on the significance of the simple correlation.
If the correlation is significant, you can go to regression and you will get good R-square.
The best way to do this is to use the stepwise regression to solve the problem of multicollinearity.
good luck
Simply you can used both test on same study. if you first have significant results of Correlations. Good Luck
Thank You all for the guidance and wonderful suggestions, i now have a strong basis to complete my work. God Bless you All
Halo dear friend,
Actually you can use any theory that you feel very relevant to your research. The best researcher always considers the other opinion for the best results. All the best.
- Badges
- Science topic
- Similar topics
- Methodology
- Research Methodology
- Research Papers