econ

Attribution Non-Commercial (BY-NC)

27 views

econ

Attribution Non-Commercial (BY-NC)

- learner autonomy and vocabulary learning strategies
- Multiple Regression
- QA Excel and Excel QM ExplsM2010
- SPSS Answers (Chapter 5)
- Rcb and Rep Measures
- bus173chap11-1
- Pricing Strategy
- Determinants of Portfolio Choice of Individual Investor
- Sas Simple Regression 2010
- A Tutrial on Correlation, Regression Analysis and Analysis of Variance
- MATH2931 Lecture 6
- Scatterplot CheckList of Questions
- Dialnet-FactorsAffectingLandUseDecisionsInThePeninsulaOfSa-5961283
- Output
- Hasil Olah
- Vidal 2012
- deepu final.docx
- Ros 3.docx
- EFFECT OF CONSUMER CHARACTERISTICS ON PURCHASE MOTIVES AND ATTITUDE TOWARDS LIFE INSURANCE: STUDY IN PUNJAB REGION
- Regression Analysis,Tools and Techniques

You are on page 1of 8

This is a version of Appendix 2 of the book Applying regression and correlation: A guide for students and researchers. This is a version of Appendix 2 of the book Applied Regression Analysis: A guide for students and researchers. It is intended to be read with the book, but should work as a standalone guide to doing regression with Excel Go Back to Jeremy Miles Homepage Books by Jeremy Miles Applying Regression Analysis Homepage Applying Regression Analysis Extras Page Appendix 2 Page In this section we will describe how to use the Excel 2000 for Windows to carry out some of the procedures we have described in the book. While this approach uses Excel 2000, I don't think that the regression procedure has changed between different versions. To do regression in Excel, you need the Analysis Toolpak add-in to be installed in Excel. This was an option when you installed Excel, but you might not have selected it. If you didn't install it, Excel will ask you for the CD, when you try to add the toolpak. Check that the add-in is installed, and added-in, by choosing Add-ins from the tools menu (as shown below).

You can now use the data analysis functions in Excel, which include multiple regression. The example that we will work through is taken from dataset 6.1b in the book "Applying regression and correlation" (if you jumped straight in here, that is what these web pages is about. Don't worry about that though, you can download the data from here .) To get to the data analysis function in Excel, you select the Tools menu, and then choose Data Analysis. This gives the following Dialog, click on Regression and then click OK.

In here, we tell Excel about the data that we would like to analyse. The first box is the input Y range. Here, we tell Excel about our dependent variable. The dependent variable must be a column, 1 cell wide, and N cells long (where N is the number of individuals that we are analysing). The the dataset we are using, the dependent variable is Anx, which is the column which goes from cell D1 to Cell D41. You can either type this information in directly as D1:D41, or you can select the appropriate data from the spreadsheet. Because we have included row 1, which includes the variable name, we are going to have to tell Excel this, by clicking on the "Labels" checkbox. The next stage is to input the independent variables. The independent variables must be a block of data, of k columns (where k is the number of independent variables) and N rows (where N is still the number of people). In the dataset we are using we have three independent variables: hassles, hassles2 and hassles3. (These represent the linear, quadratic and cubic effects of hassles - we are analysing a non-linear relationship here,) These are held in rows 1 - 41 of columns A, B and C. Again, we can type in A1:C41 or select the data from the spreadsheet - it will have the same effect. Next we tell Excel where we want the results to be written. It is best to ask for a new sheet - you don't want to accidentally overwrite some of your precious data, and have to go to all of the effort of restoring it from a backup, do you? (You do have a backup, don't you?)

We can ask fro residuals and standardised residuals to be saved - these will be new columns of numbers created in the the new spreadsheet. Two types of graphs will be drawn automatically if you ask for them. A residual plot will draw scatter plots of each independent variable on the x-axis, and the residual on the y-axis. A line fit plot will draw scatterplots of each independent variable on the x-axis, and the predicted and actual values of the dependent variable on the y axis. You cannot, as far as I have been able to determine, automatically have A scatterplot with the predicted values on the x-axis, and the residuals on the y-axis (although you can calculate these values and save them.) You can also request a normal probability plot. This appears to be a plot of the dependent variable, which is a curious thing to plot - regression analysis does not assume normal distribution of the dependent variable. The usual plot of this type would be the residuals, but this is not possible in Excel. The dialog box now looks like this:

And we get a lot of output, written to a new sheet. A note about this output - output from analysis in Excel is usually "live" that is to say, the data are linked to the output. If you change the data, you will change the output. This is not the case for this type of output in Excel. The results of the analysis are "dead" and will not change.

The Results

The results presented below have been copied from Excel to HTML, so they don't look identical. You can download the Excel file exactly as Excel wrote it for me here . Note that the data are not included in this file (it's an Excel 4.0 file, to make it usable by as many people as possible). Regression Statistics Multiple R R Square Adjusted R Square Standard Error Observations 0.807582 0.652189 0.623205 6.686129 40

The first part of the output is the regression statistics. These are standard statistics which are given by most programs. ANOVA ANOVA df Regression Residual Total 3 36 39 SS 3017.745 1609.355 4627.1 MS 1005.915 44.70432 F 22.50151 Significance F 2.22E-08

The ANOVA table comes next. This gives a test of significance of the R2. Note that Excel uses scientific notation, by default, so when it says 2.22E-08 it means, 2.22 * 10-8 . (i.e. 0.0000000222). Coefficients Coefficients Intercept HASSLES HASSLES2 HASSLES3 12.39 -0.06 0.00 0.00 Standard Error t Stat PLower value 95% 0.00 0.85 0.95 0.61 4.48 -0.75 -0.02 0.00 Upper 95% 20.29 0.62 0.02 0.00 Lower 95.0% 4.48 -0.75 -0.02 0.00 Upper 95.0% 20.29 0.62 0.02 0.00

The next stage is the coefficients. Note that here I have converted the numbers to 2 decimal places to save space). It gives the coefficient for each parameter, including the intercept (the constant). The standard errors, and the t-values follow (the t-nu value is the coefficient divided by the standard error). Next comes the p-value associated with the variable, and the confidence intervals of the parameter estimates (Excel gave these to me twice, even though I didn't ask for them.) Residuals Observation 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 Predicted Residuals ANX 12.08 11.82 15.98 29.86 28.56 34.16 12.32 12.37 27.31 15.64 -2.08 0.18 5.02 -13.86 -1.56 -4.16 -3.32 -5.37 4.69 -4.64 Standard Residuals -0.32 0.03 0.78 -2.16 -0.24 -0.65 -0.52 -0.84 0.73 -0.72

The final part of the output is the residual information. The observation in the left had column is the case number - although Excel never told us about this, it has labelled the first person Observation 1, the second Observation 2, etc. (Note that this is NOT the original row number - Observation 1 was row 2). The predicted anxiety score is the score that was predicted from the regression equation. The residual is the raw residual - that is the difference between the predicted score and the actual score on the dependent variable. The final value is the standardised residual (the residuals adjusted to ensure that they have a standard deviation of 1; they have a mean of zero already). Graphs Finally we will have a quick look at the graphs. The first graph is an example of the residual plots - it has hassles on the x-axis and the unstandardised residual on the y-axis.

The second graphs shows the predicted and actual anxiety scores plotted against hassles3.

A Note to End On

I have written this file which shows you how to do regression in Excel, but this does not mean that I think that you should be doing regression in Excel. Regression in Excel has a number of shortcomings, which include: No standardised coefficients. It can be very difficult to interpret unstandardised coefficients. You could calculate the standardised coefficients using the unstandardised coefficients, if you really wanted to. But you could have done the regression on your own, if you really wanted to. Lack of diagnostic graphs. The standard diagnostic graphs are not available in Excel, e.g. the normality plot of the residuals, the scatterplot or residuals against predicted values. Again, you can work them out, but it ain't easy. Lack of Diagnostic statistics. There are no collinearity diagnostics, which would help you to understand what was happening in the data that we analysed above (highly significant R2, but no significant parameters in the model). You could calculate the skew statistics, but not if you have more than 30 cases. Lack of features. There is no hierarchical regression, no weighting cases, etc, etc, etc...

Inflexibility. If you want to run a slightly different analysis, it is hard work, because you have to move your data around, a process which is prone to errors.

- learner autonomy and vocabulary learning strategiesUploaded byEbrahim Azimi
- Multiple RegressionUploaded byNilton de Sousa
- QA Excel and Excel QM ExplsM2010Uploaded byLeslie An
- SPSS Answers (Chapter 5)Uploaded byAbhijit Singh
- Rcb and Rep MeasuresUploaded byMuhammad Naeem Iqbal
- bus173chap11-1Uploaded byShahriar Noor
- Pricing StrategyUploaded byAlanBrett
- Determinants of Portfolio Choice of Individual InvestorUploaded byVarun Jindal
- Sas Simple Regression 2010Uploaded byngyncloud
- A Tutrial on Correlation, Regression Analysis and Analysis of VarianceUploaded byTI Journals Publishing
- MATH2931 Lecture 6Uploaded byBob
- Scatterplot CheckList of QuestionsUploaded byNgo Cici
- Dialnet-FactorsAffectingLandUseDecisionsInThePeninsulaOfSa-5961283Uploaded byTERIYAKI
- OutputUploaded byNajwa Naydirra Nanie Triana
- Hasil OlahUploaded byIrwan Cungkring
- Vidal 2012Uploaded byWilliam Rolando Miranda Zamora
- deepu final.docxUploaded bySneha Bhaip
- Ros 3.docxUploaded bySang Penggila Hujan
- EFFECT OF CONSUMER CHARACTERISTICS ON PURCHASE MOTIVES AND ATTITUDE TOWARDS LIFE INSURANCE: STUDY IN PUNJAB REGIONUploaded byAnonymous CwJeBCAXp
- Regression Analysis,Tools and TechniquesUploaded byEngr Mujahid Iqbal
- 03_LinearRegression-1Uploaded byFurqan Arshad
- CHAP 11.pptUploaded bySenelwa Anaya
- pcbtempUploaded byholubo2
- 2wayrandUploaded byAbiyyah Wiriana
- Stats 200 Problem Set 7Uploaded byPedro
- 13028-26330-1-SM.pdfUploaded bykirana
- L11 UnbalancedUploaded byTeflon Slim
- Resume Mtk Ch 8aUploaded byNurul Audina
- Brochura IBM SPSS AmosUploaded byDedi Supiyadi
- Bruno_Frontini.pdfUploaded byAngel Canales Alvarez

- How to Solve a Problem in ExcelUploaded bypalak32
- Step 3 WebsitesUploaded bypalak32
- Antibiotics Made EasyUploaded bypalak32
- Oral AntibioticsUploaded bypalak32
- Chap 007Uploaded bypalak32
- Chap 006Uploaded bypalak32
- How to Become MillioniareUploaded bypalak32
- Welcome to Class Fall 2010]Uploaded bypalak32
- sci006-lev.11Uploaded bypalak32
- Chap 010Uploaded bypalak32
- Performance_Measurements Operational ManagementUploaded bypalak32
- Introduction Operational ManagementUploaded bypalak32
- Student Ch 01Uploaded bypalak32
- Chap 015Uploaded bypalak32
- Fall 2010 Quiz (1)Uploaded bypalak32
- PSU AUA Fall 2010 Grading Explanation (1)Uploaded bypalak32
- pozgar-chapter01Uploaded bypalak32
- Assignment 3Uploaded bypalak32
- Syllabus Economic Analysis Fall 2011Uploaded bypalak32
- Ch10Uploaded bypalak32
- Student Ch 03Uploaded bypalak32
- Ch01Uploaded bypalak32
- Assignment 3Uploaded bypalak32
- Ethics Ch01 IntroUploaded bypalak32
- BU 5700 AU2 Allen Marketing Techiques Spring II May 17Uploaded bypalak32
- Student Ch 04Uploaded bypalak32
- Ethics Ch00 Intro GuideUploaded bypalak32

- SRAM_DRAM_FLASH.pdfUploaded byVishnu Vardhan Reddy
- Klipsch RW-12Uploaded byvintagekiki
- Monkey's PawUploaded byaditya123sridhar
- SERV1_01E_SystemFamilyUploaded bynyamai20008234
- Tableau Antidiabetiques Guide Des Produits 2017 EnUploaded byHayna Rose
- Lp Solver Excel GuideUploaded byKhairi Azmi
- PRMO 2017 Paper With SolutionUploaded byAnonymous UZ0mswY
- Ambulance Drone Using Raspberry Pi 3 in Emergency HealthcareUploaded byAbhishek Mishra
- To compare the effect of stretching and core strengthening exercises on Primary Dysmenohrrea in Young femalesUploaded byInternational Organization of Scientific Research (IOSR)
- PDF_Dosing_wirh_Chlorella.pdfUploaded bybrujita86
- Play GreenUploaded byhunabmc
- Bridge Procedures GuideUploaded byapi-20017870
- Prosthetic Knee SystemsUploaded byhppresario
- filipowicz2008 miRNA.pdfUploaded byChristian Moises Casas
- Some Rare Combinations in the Compendium of RhetoriusUploaded bym-imran
- basic surgical skill.pdfUploaded byEvan Anindito
- Leo CompatabilityUploaded byBrendon Peter
- David Humes Criticisms of the Cosmological ArgumentUploaded byJane Jones
- 0610_w14_ms_32Uploaded byHaider Ali
- worksheet on hydrocarbons for IIT JEEUploaded byrajput1287
- Lab 1Uploaded byDoan Thanh Thien
- Summative Test in Science 7Uploaded byJohn Perseus Lee
- Introduction to Green DesignUploaded byshatakshi_1
- Mapworld & CheatsUploaded byKetan Patel
- Octane BlendingUploaded byCarlos
- Arcelomittal –SA - Hot-dip Galvanised Cold Rolled and Hot Rolled Steel SubstrateUploaded byHen8
- Neurosurgery in PregnancyUploaded byHendrik Abraham
- Ht CatalogueENwebUploaded bylaura
- Pareto Optimal Based Evolutionary Approach for Solving Multi-Objective Facility Layout ProblemUploaded byRipon Kazi Shah Nawaz
- MARKETING PLAN FOR Tesla CompanyUploaded byacademicproffwritter