Program package SIMPLREG: SIMPLe REGression analysis This program package conducts various elementary and advanced econometric estimation and testing tasks in a simple, menu-driven, way. In particular, SIMPLREG contains modules that conduct: - Linear regression analysis, which in the time series case can be extended to models with ARMA errors or models with ARCH errors. - Instrumental variables estimation (two-stage least squares). - Limited dependent variables modelling, i.e. logit/probit analysis, Poisson regression and Binomial regression. - VAR regression and innovation response analysis, including structural VAR. - Nonlinear regression analysis, maximum likelihood estimation, and nonlinear GMM estimation. In order to use these modules, you have to buy and install Microsoft Quick-Basic 4 first. The regression, likelihood, and moment functions are specified by entering Quick-Basic code, using the Quick-Basic editor called by SIMPLREG. - Tests of linear parameter restrictions. - Test of homoskedasticity of errors against ARCH. - Bierens' integrated conditional moment test for misspecification of the functional form, and De Jong's version of this test for testing correctness of dynamic models. - Tests of the unit root hypothesis for univariate time series, i.e., the Phillips-Perron tests, the Bierens-Guo Cauchy tests of the null of (trend) stationarity against the unit root (with drift) hypothesis, Bierens' unit root tests on the basis of higher-order sample autocorrelations, the augmented Dickey-Fuller test, and Bierens' tests of the unit root with drift hypothesis against nonlinear trend stationarity. - Tests for cointegration, i.e., Johansen's cointegration tests and Bierens' nonparametric cointegration test. - Reading and merging data files, in various forms. - Searching the hard disk for suitable data files. - Viewing the current input file. - Viewing the output file. - Calculating summary statistics of the data: sample mean and standard error, minimum, maximum, effective sample size, and the quantiles in steps of 10%. - Plotting time series. - Plotting partial sums of time series. - Drawing scatter diagrams. - Plotting the time path of a bivariate time series. - Calculation and plotting of auto- and cross-correlations. - Transformations of variables: linear and multiplicative combinations, log, exp, differences, percentage changes, lags, moving averages, partial sums, dummy variables for subsets, seasonal dummy variables, indicator functions, truncation, trends, constants, and customized transformations. All the modules are menu-driven, and as self-explanatory as possible. Thus, this program package is simple in use, but advanced in contents. There is also an even more 'simple' version of SIMPLREG available, in which the advanced econometric features are hidden. This version is designed to be used in undergraduate and MA/MBA level applied econometrics courses. In particular, in this version the module for testing linear restrictions has been disabled, so that the students have to conduct the F test and LR test of parameter restrictions by hand. This version can be easily converted into the advanced one, without reinstallation, by deleting one of the files in directory \SIMPLREG. Try to figure out which one! Alternatively, you can upgrade SIMPLREG to the advanced version by correctly answering three econometric questions! The program creates a sub-directory SIMPLREG.DAT, containing the data files, the output file SIMPLREG.DAT\OUTPUT, and various auxiliary files, in the directory from where SIMPLREG is called. If you call SIMPLREG from a directory already containing the sub-directory SIMPLREG.DAT, you can continue the previous session. All non-graphical output is stored and can be printed later on, but in order to print the graphical output, one has to make screen dumps. The default format of SIMPLREG data files is: k vmis Name of variable 1 Name of variable 2 .............. Name of variable k x(1,1) x(1,2) .... x(1,k) x(2,1) x(2,2) .... x(2,k) ......................... x(n,1) x(n,2) .... x(n,k) where 'k' is the number of variables, 'vmis' is the missing value code, 'n' is the number of observations, and x() is the data matrix. Numerical data are separated by a comma or a blank. The data matrix x() may also be entered in vector form, with each data point x(t,i) on a separate line, or on one line only, as long as the order x(1,1), x(1,2),..., x(1,k),......,x(n,1), x(n,2),....,x(n,k) is preserved. The missing value code 'vmis' should be such that x(i,j) = vmis only if data point x(i,j) is a missing value, for example, set vmis = -99999. Set vmis = 0 if and only if there are no missing values. In the time series case missing values are only allowed at the beginning or the end of the time series. In the cross-section case missing values are allowed anywhere. However, if so, one can no longer write OLS residuals to the data file. The variable names may be up to 60 characters long, and may contain blanks, numbers, etc., except the characters #, &, ^, and $. A second format for SIMPLREG data files is: Name(1) Name(2) .... Name(k) x(1,1) x(1,2) .... x(1,k) x(2,1) x(2,2) .... x(2,k) ............................ x(n,1) x(n,2) .... x(n,k) where Name(i) is the name of variable i. These names should not contain blanks or commas, and should appear on the first line only, as otherwise SIMPLREG cannot determine the correct number of variables 'k'. SIMPLREG will ask for the missing value code. The third and fourth formats for SIMPLREG data files are: x(1,1) x(1,2) .... x(1,k) x(2,1) x(2,2) .... x(2,k) ............................ x(n,1) x(n,2) .... x(n,k) and its transpose: x(1,1) x(2,1) .... x(n,1) x(1,2) x(2,2) .... x(n,2) ............................ x(1,k) x(2,k) .... x(n,k) respectively. SIMPLREG will ask you for the number of variables 'k', the variable names and the missing value code 'vmis'. SIMPLREG copies the data file to file INPUT in directory SIMPLREG.DAT, in default format, which serves as a backup file. If the previous data file is of the same frequency, you may merge the old and new data file. SIMPLREG actually uses only a second copy of the (merged) data file, namely the file SIMPLREG.DAT\INPUT1. All transformations of variables are written to the latter file. In the directory SIMPLREG you will find various data files to play with. These data files are already in the default format: - NELPLOS.DAT, containing the extended Nelson-Plosser annual time series for the US, from 1860 to 1988 (extended by Peter Schotman and Herman van Dijk). - SWEDEN.DAT, containing three annual time series for Sweden from 1861 to 1988. - MDUSA1.DAT and MDUSA2.DAT, containing monthly data for the US from 1960.01 to 1987.12, and 1939.01 to 1995.01, respectively. - QDUSA.DAT, containing quartely data for the US from 1939.1 to 1995.1, and - CSDUTCH.DAT, containing a cross-section of 2000 Dutch wage earners, used by Bierens and Hartog (1988), Journal of Econometrics 38, 269-299.

Back