Is the excel sheet ok or I need econometrics tools such as SPSS and MATLAB?
Yes, I think Excel does a lot of cool stuff including correlations and multiple regression. Here in this attached link the process of using the excel (French version) to calculate the Pearson correlation coefficient.
https://help.xlstat.com/s/article/coefficient-de-correlation-de-pearson-dans-excel?language=fr
About multiple and simple regressions, Excel also do them. For multiple linear regression:
Once XLSTAT has been launched, choose the XLSTAT / Modeling / Linear regression command. Once the button is clicked, the dialog box corresponding to the regression appears. In the General tab, you can then select the data on the Excel sheet. Here are corresponding links (French versions) both for simple linear regression and multiple regression:
https://help.xlstat.com/s/article/regression-lineaire-simple-dans-excel?language=fr
http://eric.univ-lyon2.fr/~ricco/tanagra/fichiers/fr_Tanagra_Regression_Excel.pdf
Of course, other softwares are interesting for correlations and multiple regressions like STATISTICA, Minitab, SAS and others.
Dear Kimosop,
Excel can do but has some limitations as it does have several functionalities to perform certain tests/estimations - depending on the relationship(s) among variables you intend to investigate and the correct procedure and suitable estimation technique that may have to applied in the estimation process.
In addition to the softwares you mentioned above, I would further recommend that you can use Stata, EViews, Minitab, Microfit, R, SAS, Gretl, etc.
Yes, there is a function for correlation already installed by default and you can perform also multiple regression.
https://www.youtube.com/watch?v=Q5JlRmmHzsg&ab_channel=Dr.ToddGrande
http://cameron.econ.ucdavis.edu/excel/ex61multipleregression.html
Yes, I think Excel does a lot of cool stuff including correlations and multiple regression. Here in this attached link the process of using the excel (French version) to calculate the Pearson correlation coefficient.
https://help.xlstat.com/s/article/coefficient-de-correlation-de-pearson-dans-excel?language=fr
About multiple and simple regressions, Excel also do them. For multiple linear regression:
Once XLSTAT has been launched, choose the XLSTAT / Modeling / Linear regression command. Once the button is clicked, the dialog box corresponding to the regression appears. In the General tab, you can then select the data on the Excel sheet. Here are corresponding links (French versions) both for simple linear regression and multiple regression:
https://help.xlstat.com/s/article/regression-lineaire-simple-dans-excel?language=fr
http://eric.univ-lyon2.fr/~ricco/tanagra/fichiers/fr_Tanagra_Regression_Excel.pdf
Of course, other softwares are interesting for correlations and multiple regressions like STATISTICA, Minitab, SAS and others.
Yes.
You can do it using Excel. However, there are some other statistical tools most adjusted than Excel such as Statistica and SPSS, for instance. These tools are a litle more difficult however the obtained results will be more robust and comprehensive.
Best regards
JSCamara
Yes but certain statistical tests may not be available as shared earlier in the normal excel statistical package.
However, you can incorporate this into Excel, like Real Statistics.
http://www.real-statistics.com/free-download/real-statistics-resource-pack/
Many other things can be done with VBA in Excel including those in other software.
Yes but i recommend you to use Mini Tab or SPSS to get robust results
It depends on 2 things. 1. What's your research question ? and 2 Are you just doing your thesis for the hell of it or do you intend to make research your career? Now most research questions do not lend themselves to Excel and further Excel has a history of getting the computations wrong(often a poor choice of algorithms). To any that are going into research in a serious way( say methods beyond simple regression and correlation coefficients) take a serious look at these 2 links:
1. Computer Age Statistical Inference: Algorithms, Evidence, and Data Science | Bradley Efron, Trevor Hastie | download (b-ok.cc) and
2. R for everyone: advanced analytics and graphics | Lander, Jared P | download (b-ok.cc)
You can compete with anybody with these tools. Best of luck with your research, David Booth
Spss&matkab program can do it,although Excel can do these jobs including pearson correlation&multiple regression.
I agree with David. Correlation analysis, yes Excel can do but with regards to multiple regression analysis, it depends on the depth of the analysis
Yes (using the function LINEST), unless you have missing values (EXCEL can not handle missing values in multiple regression). EXCEL is generally very bad in handling missing values. Also, it gives completely wrong results when the model if over-specified and when there is a very high correlation between predictors. Further, Excel is not explicitely separating data, analysis and results. This is considered a bad practice in science. If you like to follow "good scientific practice" in your work as far as possible, this should be a reason for better not using EXCEL in your thesis.
Apart from these problems (that can be avoided when taking care), EXCEL is really making everything difficult. It is difficult in EXCEL manage your model (you must create the design matrix manually; you have to use "matrix formulae", what can be tricky). It is difficult to get proper diagnostic plots. It is difficult to create good scientific diagrams with EXCEL in general. And it is difficult to get relevant statistics like confidence and predcition bands etc. Nothing is really impossible. But almost everything is difficult.
If you do that only for your present work, knowing that you will never do data analysis again, it may be ok (but still difficult and error-prone). However, as David Eugene Booth stated above, if the odds are that you will do data analysis more often in your career, it's highly highly recommended to use some more scientific software.
I recommend you the use of this Excel Addin: https://www.real-statistics.com/
With it you can perform Multiple Lineal Regression.
In my opinion there is no better or worse software to do statistical analysis. If Excel is capable of doing what you want, it is good software. You can also use the Visual Basic for Applications programming language if you want to do more complex things. An advantage over other calculation software is that it is cheap and that it allows the information to be organized spatially. I use it on a daily basis for graphing and routine statistical analysis. It is true that it does not allow to do certain more specific analyzes, that maybe are implemented in Python or in R.
Todas son útiles para lo que necesitas emplear, pero está supeditada a los niveles de conocimientos que tengas con cada una de estas herramientas.
Good afternoon.
If multiple regression is possible with EXCEL, but if you can do it with SPSS or Matlab, I suggest you do it, due to the versatility and indicators that you can obtain for the testing of the hypotheses.
- Badges
- Science topic