Comparing free
statistical software
For data sets with no
missing values
Click here
to return to the free software page
This page shows some output
from the programs listed below. The
output is
correlations and regression. I did this in November 2006 using the most
recent versions of the software at that time. I used a dataset with 4
variables, a subset from PD-Plus, available on my data page. I
updated this in November 2008 with OpenStat and PSPP
Easyreg
http://econ.la.psu.edu/~hbierens/EASYREG.HTM
Epidata
http://www.epidata.dk/
Instat http://www.ssc.rdg.ac.uk/software/instat/instat.html
OpenStat http://www.statpages.org/miller/openstat/
WinIDAMS http://portal.unesco.org/ci/en/ev.php-URL_ID=2070&URL_DO=DO_TOPIC&URL_SECTION=201.html
MicrOsiris http://www.microsiris.com/
Epi Info
2000 Windows
http://www.cdc.gov/epiinfo/index.htm
PSPP http://www.gnu.org/software/pspp/
Using a data set with
all
cases (no missing values):
http://gsociology.icaap.org/methods/fourvars.csv
1. All programs read .csv
files, except epi info, which imports excell files, among other formats.
2. When using MicrOsiris,
a.
import the .csv file, then call up commands.
b.
for blanks, Microsiris assigns 1.5 and 1.6 billion, but automatically
recognises these values as missing.
c.
the data dictionary shows 0 decimal places, but if the data actually
have decimal places, like 1.23, the number is read as 1.23, with the
decimal place. The data dictionary shows how many decimal places
are implied, if there isn't
one.
3. When
using WinIDAMS,
all values of each variable should have the same number of decimal
places. So you need to open the file above to excel, format each
variable to, say, 2 decimal places. Also, WinIDAMS can't handle
variables with more than 10 digits.
4. When using OpenStat,
I had to remove the first line of the data set, that had variable
names. I'm not sure whether my data are especially strange. So
for openstat, I used this data set
http://gsociology.icaap.org/methods/fourvars_nonames.txt
Stat4U (previous version of OpenStat) seems to have a problem when
there is too much variation in a
variable, like population varying from a billion to a thousand.
Haven't checked this out with openstat yet.
5. When using epistat,
regression make the dependent variable the
first in the list.
6. When using PSPP, I
had to create another dataset. For some reason, it did not
recognize the last variable, so I added a dummy at the end.
http://gsociology.icaap.org/methods/fourvars_v2.csv
7. Correlation:
- All
programs give exactly the same correlation
coefficients, except OpenStat gives coefficients that seem to be
slightly different. I'm trying to find out why.
- OpenStat also gives statistical significance levels for the
correlations. WinIDAMS gives t-tests. MicrOsiris also gives
significance levels, t-tests.
- Epi Info doesn't do simple correlations. You have to do
regression with just two variables to get the correlation.
- PSPP does not yet
have correlation.
8. Regression:
all programs
give the same results for basic regression
and some same results for backward/forward stepwise.
- Epidata, Epi Info, Instat+ and PSPP don't
allow choices of what type of output.
- Output from Epidata, Epi Info, Instat+ and PSPP are the same.
Also the same as
WinIDAMS stepwise and an Easyreg regression.
- If you use any program to
run a full model or best fit with all variables, then eliminate the
variable that isn't significant, and then rerun the regression.
- When I use WinIDAMS
stepwise and WinIDAMS decending, I get the same results.
- I can't match OpenStat
backwards regression with any other output. I don't know why.
- MicrOsiris has
stepwise,
but I haven't figured it out yet.
Misc notes:
1. Epi info doesn't have a menu command for getting means of multiple
variables, can only seem to get means for one variable at a time.
Return
to top
This is a sample of papers that use these packages. Many sites link to
them as well. I just list some places that use EasyReg, Stat4U,
EpiData, Instat, Irristat, MicrOrsiris, as the other programs
(WinIDAMS, EpiInfo) are from major institutions (UNESCO, CDC) so are
pretty well used.
EasyReg
Spatial and Temporal Transferability of Trip Generation Demand Models
in Israel, A Cotrus, J Prashker, Y Shiftan
Journal of Transportation and Statistics,
http://www.bts.gov/publications/journal_of_transportation_and_statistics/volume_08_number_01/html/paper_04/
mentions using Easyreg when calculating R squared, in table 10.
On the relationship between the market risk premium and the risk-free
interest rate. Confidence W. Amadi
A Journal of Applied Topics in Business and Economics, 2004
http://www.westga.edu/~bquest/2004/relationship.htm
Cushman, David O., (2003) "Further evidence on the size and power of
the Bierens and Johansen cointegration
procedures." Economics Bulletin, Vol. 3, No. 25 pp. 1−7
http://economicsbulletin.vanderbilt.edu/Abstract.asp?PaperID=EB-03C30001
Predicting Financial Time Series by Genetic Programming with
Trigonometric Functions and High-Order Statistics
R. Schwaerzel and T Bylander, http://www.cs.utsa.edu/~rschwaer/AppliedSoftComputing.pdf
Assessing the Impact of the September 11 Terrorist Attacks on U.S.
Airline Demand
H Ito and D Lee, http://www.brown.edu/Departments/Economics/Papers/2003/2003-16_paper.pdf
Epidata
Epidata is listed in a CDC MMWR report
http://www.cdc.gov/mmwr/preview/mmwrhtml/su5501a6.htm
and here
Houston JM, Martin M, Williams JE, Hill RL. The Annual African American
Conference on Diabetes: evolving program evaluation with evolving
program implementation. Prev Chronic Dis [serial online] 2006 Jan [date
cited]. Available from: URL: http://www.cdc.gov/pcd/issues/2006/jan/05_0119.htm
Some brief guides here http://www.son.wisc.edu/RDSU/datamgt.html
Epi Info
Seroprevalence of hepatitis C and associated risk factors in urban
areas of Antananarivo, Madagascar
Charles E Ramarokoto and others. BMC Infectious Diseases 2008,
8:25
http://www.biomedcentral.com/1471-2334/8/25/
Rahav G, Gabbay R, Ornoy A, Shechtman S, Arnon J, Diav-Citrini O.
Primary versus nonprimary cytomegalovirus infection during pregnancy,
Israel. Emerg Infect Dis [serial on the Internet]. 2007 Nov [May 15,
2009]. Available from http://www.cdc.gov/EID/content/13/11/1791.htm
Chan P-C, Huang L-M, Wu Y-C, Yang H-L, Chang I-S, Lu C-Y, et al.
Tuberculosis in children and adolescents, Taiwan, 1996–2003. Emerg
Infect Dis [serial on the Internet]. 2007 Sep. Available from
http://www.cdc.gov/EID/content/13/9/1361.htm
ME Gyasi, WMK Amoaku, and MA Adjuik. Epidemiology of Hospitalized
Ocular Injuries in the Upper East Region of Ghana. Ghana Med J. 2007
December; 41(4): 171–175. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2350113
Nocardial infections: report of 22 cases
Maria Bernadete F. Chedid; Marcio F. Chedid; Nelson S. Porto;
Cecília B. Severo; Luiz Carlos Severo
Rev. Inst. Med. trop. S. Paulo vol.49 no.4 São Paulo Jul./Aug.
2007
http://www.scielo.br/scielo.php?pid=S0036-46652007000400009&script=sci_arttext&tlng=en
Intestinal parasitosis and nutritional status in schoolchildren of
Sahar district, Yemen
Y.A. Raja’a and J.S. Mubarak
Eastern Mediterranean Health Journal, Vol. 12 (Supplement 2), 2006
S189
http://www.emro.who.int/Publications/EMHJ/12_S2/article18.htm
Instat
Return
to top
The relationship between pyrethrins and the yellow pigmentation in
pyrethrum flowers. Wenwa A. Odinga and Charles A. Angedu. 2003. African
Journal of Science and Technology (AJST), Science and Engineering
Series Vol. 4, No. 2, pp. 116-123. http://www.ansti.org/volume/Odinga-final.pdf
This journal is from the African Network of Scientific and
Technological Institutions, part of UNESCO and UNDP.
A STAT5 modifier locus on murine chromosome 7 modulates engraftment of
hematopoietic stem cells during steady-state hematopoiesis
Christine Couldrey, Heath L. Bradley, and Kevin D. Bunting.
Blood, 15 February 2005, Vol. 105, No. 4, pp. 1476-1483. http://bloodjournal.hematologylibrary.org/cgi/content/full/105/4/1476
Metabotropic Glutamate Receptors and Dopamine Receptors Cooperate to
Enhance Extracellular Signal-Regulated Kinase Phosphorylation in
Striatal Neurons. Voulalas et al. The Journal of Neuroscience, April
13, 2005, 25(15):3763-3773. http://www.jneurosci.org/cgi/content/full/25/15/3763
Association of the XRCC1 gene polymorphisms with cancer risk in Turkish
breast cancer patients. Deligezer and Dalay. Experimental and molecular
medicine, Vol. 36, No. 6, 572-575, December 2004. http://www.e-emm.org/article/article_files/EMM036-06-10.pdf
This journal is published by The Korean Society of Medical
Biochemistry and Molecular Biology.
Irristat
I haven't used it yet, but IRRISTAT is used here
FAO Plant Production and Protection Paper No. 174, Rome, 2003, Genotype
x environment interactions. Challenges and opportunities for plant
breeding and cultivar recommendations, listed here http://www.fao.org/catalog/bullettin/07_03.htm
The report is here http://www.fao.org/DOCREP/005/Y4391E/y4391e00.htm
and Irristat is listed in the forward and the following pages.
http://www.fao.org/docrep/005/Y4391E/y4391e0b.htm,
http://www.fao.org/DOCREP/005/Y4391E/y4391e08.htm
Yield and Soil Nutrient Changes in a Long-Term Rice-Wheat Rotation in
India
A. L. Bhandari, J. K. Ladha, H. Pathak, A. T. Padre, D. Dawe and R. K.
Gupta
Soil Science Society of America Journal 66:162-170 (2002)
http://soil.scijournals.org/cgi/content/full/66/1/162
Effects of Residue Decomposition on Productivity and Soil Fertility in
Rice–Wheat Rotation
Yadvinder-Singh, Bijay-Singh, J. K. Ladha, C. S. Khind, T. S. Khera and
C. S. Bueno
Soil Sci. Soc. Am. J. 68:854-864 (2004).
http://soil.scijournals.org/cgi/content/full/68/3/854
An agro-economic analysis of grain production in Estonia after its
transition to market economy
N. Vasiliev, A. Astover, H. Roostalu, E. Matveev
Agronomy Research, Vol 4 (1)
http://www.eau.ee/~agronomy/
Analysis of Magnaporthe grisea population structure in Côte
d’Ivoire as a prerequisite for the deployment of varieties with durable
blast resistance.
Y. Séré et al, in Department for International
Development-Crop Protection Programme (DFID-CPP) funded project - Rice
blast in West Africa workshop papers
http://www2.warwick.ac.uk/fac/sci/whri/about/staff/sprasad/
MicrOsiris
History and Potential of Binary Segmentation for Exploratory Data
Analysis. James N. Morgan.
Journal of Data Science, v.3, no.2, 123-136
http://proj1.sinica.edu.tw/~jds/A198.html
Part of WinIDAMS has sources from MicrOsiris
http://www.unesco.org/webworld/idams/newsletter_sep.html
CDC mentions that one state data system was converted to be available
for MicrOsiris
http://ftp.cdc.gov/pub/Software/RegistryPlus/Inventory_Assessment/Assessment2005.xls
and here http://www.cdc.gov/cancer/npcr/npcrpdfs/assessment2005.pdf
The decision tree is mentioned here
http://onlineacademics.org/CA517/Practitioner/Unit5CorrelationPractitioner.ppt
and here
http://www.thoracic.org/sections/clinical-information/best-of-the-web/pages/research/choosing-the-best-statistical-test.html
as an important site
OpenStat
NMDA-induced Seizure Intensity is Enhanced in COX-2 Deficient Mice
Christopher D. Toscano, Philip J. Kingsley, Lawrence J. Marnett, and
Francesca Bosetti
Neurotoxicology. 2008 November; 29(6): 1114–1120.
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2587528
Future Salary and US Residency Fill Rate Revisited
Mark Ebell. Research letter in JAMA, September 10, 2008—Vol 300,
No. 10, p1131-1132
http://jama.ama-assn.org/cgi/reprint/300/10/1131
Pulido Ximena Carolina, Pérez Gerardo, Vallejo Gustavo Adolfo.
Preliminary characterization of a Rhodnius prolixus hemolymph
trypanolytic protein, this being a determinant of Trypanosoma rangeli
KP1(+) and KP1(-) subpopulations' vectorial ability. Mem. Inst. Oswaldo
Cruz [serial on the Internet]. 2008 Mar [cited
2009 May 28] ; 103(2): 172-179. Available from: http://www.scielo.br/scielo.php?script=sci_arttext&pid=S0074-02762008000200008&lng=en
. doi: 10.1590/S0074-02762008000200008.
Differential gene expression patterns in cyclooxygenase-1 and
cyclooxygenase-2 deficient mouse brain
Christopher D Toscano, Vinaykumar V Prabhu, Robert Langenbach, Kevin G
Becker, and Francesca Bosetti
Genome Biol. 2007; 8(1): R14.
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1839133
Shiga Toxin-Mediated Hemolytic Uremic Syndrome: Time to Change the
Diagnostic Paradigm?
Martina Bielaszewska, Robin Köck, Alexander W. Friedrich, Christof
von Eiff, Lothar B. Zimmerhackl, Helge Karch, and Alexander Mellmann
PLoS ONE. 2007; 2(10): e1024
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1995754
Cytolethal Distending Toxin from Shiga Toxin-Producing Escherichia coli
O157 Causes Irreversible G2/M Arrest, Inhibition of Proliferation, and
Death of Human Endothelial Cells. Bielaszewska et al. Infection and
immunity, Jan. 2005, p. 552–562 Vol. 73, No. 1
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=538959
and
Chromosomal Dynamism in Progeny of Outbreak-Related Sorbitol-Fermenting
Enterohemorrhagic Escherichia coli O157:NM, Bielaszewska et al, Appl
Environ Microbiol. 2006 March; 72(3): 1900–1909. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1393231
Both journals are from the American Society for Microbiology, and both
use OpenStat.
Endothelium histological integrity after skeletonized dissection of the
left internal mammary artery with ultrasonic scalpel. Cañadas et
al, Interact CardioVasc Thorac Surg 2005;4:160-162. http://icvts.ctsnetjournals.org/cgi/content/full/4/3/160
from the European Association of Cardio-Thoracic Surgery.
Uses OpenStat.
Are interleukin-16 and thrombopoietin new tools for the in vitro
generation of dendritic cells? Bella et al, Blood, 15 December 2004,
Vol. 104, No. 13, pp. 4020-4028. http://bloodjournal.hematologylibrary.org/cgi/content/full/104/13/4020
A journal of the American Society of Hematology. Uses OpenStat.
Reviews which include OpenStat
Stat4U has a nice brief mention in this article "Software you can use",
in the Sept 2006 Newsletter of the Society for Judgment and Decision
Making, http://www.sjdm.org/content/newsletters
UTILIZACIÓN DE SOFTWARE DE LIBRE ACCESO PARA LA ENSEÑANZA
DE ESTADÍSTICA Y PSICOMETRÍA
Leonardo Adrián Medrano http://psicologia.udg.es/revista/publicacions/04/Cast/01(4)_Cast.pdf
(in Spanish)
PAST –
Palaeontological Statistics
http://folk.uio.no/ohammer/past/
(I haven't tried this out yet, but this seems like a well used program)
Genetic variation in the threatened medicinal tree prusus africana in
Cameroon and Kenya. Alice Muchugi, Ard G. Lengkeek, Caroline A.C.
Agufa, Geoffrey M. Muluvi, Eliud N.M. Njagi and Ian K. Dawson. FAO
International workshop, 5-7 March 2005, Turin, Italy.
http://www.fao.org/biotech/torino05.htm
Is appropriate appropriate? An investigation of interpersonal semantic
stability
H.P.L. Molloy Temple University Japan
Proceedings of the 2nd Annual JALT Pan-SIG Conference. May
10-11, 2003. Kyoto, Japan: Kyoto Institute of Technology.
http://jalt.org/pansig/2003/HTML/Molloy.htm
Seriation in Paleontological Data Using Markov Chain Monte Carlo Methods
Kai Puolamäki, Mikael Fortelius, and Heikki Mannila
Computational Biology, 2006 February; 2(2): e6
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1361357
Lipid composition of sesame seeds (Sesamum indicum L.) using
multivariate analysis
Auristela Malavé Acuña y Jesús Rafael
Méndez Natera
Bioline International, Revista Científica UDO Agrícola
Vol. 5, Núm. 1, 2005, pp. 48-53
http://www.bioline.org.br/request?cg05006
WinIdams
High Differentiation among Eight Villages in a Secluded Area of
Sardinia Revealed by Genome-Wide High Density SNPs Analysis
Giorgio Pistis and a lot of others....
PLoS ONE. 2009; 4(2): e4654.
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2646134
Computational Modeling of Substitution Effect on HIV–1 Non–Nucleoside
Reverse Transcriptase Inhibitors with Kier–Hall Electrotopological
State (E–state) Indices
Nitin S. Sapre,1 Nilanjana Pancholi,1 and Swagata Gupta
Internet Electronic Journal of Molecular Design, March 2008, Volume 7,
Number 3, Pages 55–67
http://biochempress.com/Files/iejmd_2008_7_0055.pdf
Multiple classification analysis in trip production models
Cristian Angelo Guevaraa, and Alan Thomas
Transport Policy, Volume 14, Issue 6, November 2007, Pages 514-522
http://crguevar.googlepages.com/Guevara_Thomas_JTP.pdf
(cites one of the manuals)
Counting Clusters Using R-NN Curves
Rajarshi Guha, Debojyoti Dutta, David J. Wild, and Ting Chen
J Chem Inf Model. 2007; 47(4): 1308–1318.
http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=2543137
Enzyme Activity Fingerprinting with Substrate Cocktails
Jean-Philippe Goddard and Jean-Louis Reymond
J. AM. CHEM. SOC. 2004, 126, 11116-11117
http://www.dcb-server.unibe.ch/groups/reymond/publications/95.pdf
Return
to top
*****************
EPIDATA
*****************
free00 clit
gdpcap00 imr2000
free00
1.000
clit
-0.425 1.000
gdpcap00
-0.483 0.416
1.000
imr2000
0.524 -0.758
-0.519 1.000
*****************
INSTAT
PLUS
*****************
imr2000 gdpcap0
clit free00
imr2000 1.0000
gdpcap0
-0.5190 1.0000
clit
-0.7580
0.4158 1.0000
free00
0.5236 -0.4827
-0.4252 1.0000
*****************
OpenStat
*****************
Correlations
IMR2000 GDPCAP
C-lit free
IMR2000
1.000 -0.529
-0.757 0.520
GDPCAP
-0.529
1.000 0.420
-0.482
C-lit
-0.757
0.420 1.000
-0.421
free
0.520 -0.482
-0.421 1.000
No. of valid cases = 171
*****************
WINIDAMS
*****************
VAR
2
3 4
3
gdpcap00
3 -0.5190
4
clit
4 -0.7580 0.4158
5
free00
5 0.5236 -0.4827 -0.4252
*****************
EASYREG
*****************
Sample correlation matrix
1 -0.5189745339 -0.7579766302 0.5235762816
-0.5189745339
1 0.4158181979 -0.4827020603
-0.7579766302
0.4158181979
1 -0.4252419584
0.5235762816 -0.4827020603
-0.4252419584
*****************
MicrOsiris
*****************
**CORRELATIONS**
V2
V3 V4
imr2000 gdpcap00 clit
gdpcap00 V3 -0.5190
clit
V4 -0.7580 0.4158
free00 V5
0.5236 -0.4827 -0.4252
Return
to top
Using literacy rate, gdp per capita and infant mortality rate to
predict freedom.
*****************
EPIDATA
*****************
free00
clit
gdpcap00 imr2000
free00
1.000
clit
-0.425 1.000
gdpcap00
-0.483 0.416
1.000
imr2000
0.524 -0.758
-0.519 1.000
Source
Sum Sq Mean Sq
df Number of
obs 172
Model
228.49 76.16
3 F(3,168)
28.36
Residual
451.18 2.69 168
Prob > F 0.00
Total
679.67 3.97
171
R-squared 0.34
Root
MSE 1.64
Variable
Beta LCL UCL
SE
t P>|t|
imr2000
0.02 0.01 0.03
0.01 3.27 0.00
gdpcap00
-0.05 -0.08 -0.03
0.01 -3.90 0.00
clit
-0.01 -0.02
0.01 0.01 -0.53 0.59
Intercept
3.59 1.70 5.49
0.96 3.75 0.00
Total N = 172 Included: N= 172
****
JUST IMR AND
GDPCAP
Source
Sum Sq Mean Sq
df Number of
obs 172
Model
227.72 113.86
2 F(2,169)
42.58
Residual
451.94 2.67 169
Prob > F 0.00
Total
679.67 3.97
171
R-squared 0.34
Root
MSE 1.64
Variable
Beta LCL UCL
SE t P>|t|
imr2000
0.02 0.01 0.03
0.00 5.09 0.00
gdpcap00
-0.05 -0.08 -0.03
0.01 -3.93 0.00
Intercept
3.10 2.59 3.61
0.26 11.98 0.00
Total N = 172 Included: N= 172
*****************
INSTAT
PLUS
*****************
imr2000 gdpcap0
clit free00
imr2000 1.0000
gdpcap0
-0.5190 1.0000
clit
-0.7580 0.4158 1.0000
free00
0.5236 -0.4827
-0.4252 1.0000
ANOVA for regression of free00
on imr2000 gdpcap0 clit
-------------------------------------------------------------------
Source
df
SS
MS F value
Prob>F
-------------------------------------------------------------------
Regression
3
228.492
76.164
28.36 0.0000
Residual
168
451.176 2.6856
-------------------------------------------------------------------
Total
171 679.667
-------------------------------------------------------------------
R-squared = 0.3362
(adjusted = 0.3243)
*****************
OpenSTAT
*****************
Return
to top
Correlations
Product-Moment Correlations Matrix
Variables IMR2000
GDPCAP
C-lit free
IMR2000
1.000
-0.529
-0.757 0.520
GDPCAP
-0.529
1.000
0.420 -0.482
C-lit
-0.757
0.420
1.000 -0.421
free
0.520
-0.482
-0.421 1.000
BEST FIT REGRESSION
Variables entered in step 2
1 IMR2000
2 GDPCAP
Squared Multiple Correlation = 0.3298
Dependent variable = free
ANOVA for Regression Effects :
SOURCE
df
SS
MS
F
Prob
Regression 2
221.8869
110.9434
41.3269 0.0000
Residual 168
451.0020 2.6845
Total
170 672.8889
Variables in the equation
VARIABLE
b s.e. b
Beta t prob. t
IMR2000 0.01860 0.0038
0.3683 4.949 0.0000
GDPCAP -0.05640 0.0146 -0.2870 -3.857
0.0002
(Intercept) 3.13079
Increase in squared R for this step = 0.059349
F = 14.8761 with D.F. 1 and 168 with Probability = 0.0002
----------------------------------------------------------
Last variable added failed entry test. Job ended.
----------------------------------------------------------
OpenStat
BACKWARD STEPWISE
Backward Stepwise Multiple Regression by Bill Miller
----------------- STEP 3 ------------------
Determinant of correlation matrix = 0.7296
SOURCE DF
SS
MS
F Prob.>F
Regression 1 181.951
181.951 62.635 0.000
Residual 169 490.937
2.905
Total 170 672.889
Dependent Variable: free
R
R2
F Prob.>F DF1 DF2
0.520 0.270
62.635 0.000 1 169
Adjusted R Squared = 0.266
Std. Error of Estimate = 1.704
Variable
Beta
B Std.Error
t
Prob.>t VIF TOL
IMR2000
0.520 0.026
0.003 7.914
0.000 1.000 1.000
Constant = 2.426
Partial Correlations
Variables IMR2000
0.520
-----------------------------------------
OpenStat
FORWARD
STEPWISE
-----------------------------------------
-------------FINAL STEP-----------
SOURCE DF
SS
MS
F Prob.>F
Regression 2 221.887
110.943 41.327 0.000
Residual 168 451.002
2.685
Total 170 672.889
Dependent Variable: free
R
R2
F Prob.>F DF1 DF2
0.574 0.330
41.327 0.000 2 168
Adjusted R Squared = 0.322
Std. Error of Estimate = 1.638
Variable
Beta
B Std.Error
t
Prob.>t VIF TOL
IMR2000
0.368 0.019
0.004 4.949
0.000 1.388 0.721
GDPCAP -0.287
-0.056 0.015
-3.857 0.000
1.388 0.721
Constant = 3.131
*****************
WINIDAMS
STEPWISE (I think this
is best fit)
*****************
Return
to top
F-level
0.286
T-level
0.535
Standard error of
estimate
1.639
F ratio for the
regression
28.360
Multiple correlation
coefficient
0.57981
adjusted 0.56950
Fraction of explained variance (RSQD)
0.33618
adjusted 0.32433
Determinant of the correlation matrix
0.31037
Residual degrees of freedom
(N-p-1) 168
Constant
term
3.5943
Partial
Var.
no.
B Sigma(B)
Beta Sigma(Beta)
RSQD Marg RSQD T-ratio Cov.
ratio Variable name
2
0.0170 0.0052
0.3355 0.1026
0.0598 0.0422
3.2692 0.6247
imr2000
3
-0.0528 0.0135
-0.2872 0.0736
0.0831 0.0602
3.9018 0.2705
gdpcap00
4
-0.0050 0.0094
-0.0516 0.0964
0.0017 0.0011
0.5346 0.5752
clit
**************** Listing of
marginal R-squares for all potential predictors ***
Step
no. Var. no. Variable
name
Marg rsqd Categorical variables (all
codes) Previously
in (*)
Marg RSQD T-ratio
3
2
imr2000
0.0422
*
3
3
gdpcap00
0.0602
*
3
4
clit
0.0011
*
*****************
WINIDAMS
Decending
*****************
Standard error of
estimate
1.639
F ratio for the
regression
28.360
Multiple correlation
coefficient
0.57981
adjusted 0.56950
Fraction of explained variance (RSQD)
0.33618
adjusted 0.32433
Determinant of the correlation matrix
0.31037
Residual degrees of freedom
(N-p-1) 168
Constant
term
3.5943
Partial
Var.
no.
B Sigma(B)
Beta Sigma(Beta)
RSQD Marg RSQD T-ratio Cov.
ratio Variable name
2
0.0170 0.0052
0.3355 0.1026
0.0598 0.0422
3.2692 0.6247
imr2000
3
-0.0528 0.0135
-0.2872 0.0736
0.0831 0.0602
3.9018 0.2705
gdpcap00
4
-0.0050 0.0094
-0.0516 0.0964
0.0017 0.0011
0.5346 0.5752
clit
**************** Listing of
marginal R-squares for all potential predictors ***
Step
no. Var. no. Variable
name
Marg rsqd Categorical variables (all
codes) Previously
in (*)
Marg RSQD T-ratio
0
2
imr2000
0.0422
*
0
3
gdpcap00
0.0602
*
0
4
clit
0.0011
*
**************
Easyreg
**************
Return
to top
X variables:
X(1) = imr2000
X(2) = gdpcap00
X(3) = clit
X(4) = 1
OLS estimation results
Parameters
Estimate t-value H.C. t-value
(S.E.) (H.C. S.E.)
[p-value] [H.C. p-value]
b(1)
0.01698
3.269 3.485
(0.00519) (0.00487)
[0.00108] [0.00049]
b(2)
-0.05281
-3.902 -5.441
(0.01353) (0.00971)
[0.00010] [0.00000]
b(3)
-0.00503
-0.535 -0.554
(0.00941) (0.00909)
[0.59293] [0.57960]
b(4)
3.59431
3.749 3.889
(0.95867) (0.92417)
[0.00018] [0.00010]
Effective sample size
(n):
172
Variance of the
residuals:
2.685569
Standard error of the residuals
(SER): 1.638771
Residual sum of squares
(RSS):
451.175601
(Also called SSR = Sum of Squared
Residuals)
Total sum of squares
(TSS):
679.667151
R-square:
0.3362
Adjusted
R-square:
0.3243
Overall F test: F(3,168) = 28.36
p-value = 0.00000
******
EASYREG
JUST USING IMR AND GDPCAP
X variables:
X(1) = imr2000
X(2) = gdpcap00
X(3) = 1
OLS estimation results
Parameters
Estimate t-value H.C. t-value
(S.E.) (H.C. S.E.)
[p-value] [H.C. p-value]
b(1)
0.01891
5.093 5.566
(0.00371) (0.00340)
[0.00000] [0.00000]
b(2)
-0.05310
-3.935 -5.505
(0.01349) (0.00965)
[0.00008] [0.00000]
b(3)
3.10092
11.982 12.041
(0.25881) (0.25753)
[0.00000] [0.00000]
Effective sample size
(n):
172
Variance of the
residuals:
2.67422
Standard error of the residuals
(SER): 1.635304
Residual sum of squares
(RSS):
451.94311
(Also called SSR = Sum of Squared
Residuals)
Total sum of squares
(TSS):
679.667151
R-square:
0.3351
Adjusted
R-square:
0.3272
Overall F test: F(2,169) = 42.58
p-value = 0.00000
*****************
MicrOsiris
*****************
Return
to top
****************
(this is default, best fit.)
****************
Total case
count: 172
STANDARD REGRESSION
THE DEPENDENT VARIABLE IS V:
free00
STANDARD
ERROR OF
ESTIMATE
1.64
F-RATIO
FOR THE
REGRESSION
28.360 PROBABILITY 0.00
MULTIPLE
CORRELATION COEFFICIENT
0.5798 ADJUSTED 0.5695
FRACTION
OF EXPLAINED
VARIANCE
0.3362 ADJUSTED 0.3243
DETERMINANT OF THE CORRELATION MATRIX 0.31037
RESIDUAL
DEGREES OF FREEDOM (N-K-1) 168
CONSTANT
TERM
3.5943
STD. ERROR 0.958673
VARIABLE
NAME
B
SIGMA(B)
BETA SIGMA(BETA)
V2
imr2000
0.16977E-01 0.51931E-02
0.33546 0.10261
V3
gdpcap00
-0.52809E-01 0.13534E-01 -0.28717
0.73597E-01
V4
clit
-0.50331E-02 0.94148E-02 -0.51560E-01 0.96447E-01
PARTIAL PART
MARGINAL
COVARIANCE
VARIABLE
NAME
R R
RSQD T-RATIO(PROB) RATIO
V2
imr2000
0.245 0.205 0.0422 3.2692 (.002)
0.624
V3
gdpcap00
-0.288 0.245 0.0602 3.9018 (.000)
0.270
V4
clit
-0.041 0.034 0.0011 0.5346 (.600)
0.575
*****************
Epi
Info
*****************
Return
to top
Linear Regression
| Variable |
Coefficient |
Std Error |
F-test |
P-Value |
| clit |
-0.005 |
0.009 |
0.2858 |
0.593642 |
| gdpcap00 |
-0.053 |
0.014 |
15.2243 |
0.000138 |
| imr2000 |
0.017 |
0.005 |
10.6876 |
0.001310 |
| CONSTANT |
3.594 |
0.959 |
14.0569 |
0.000244 |
| Correlation Coefficient: r^2= |
0.34 |
| Source |
df |
Sum of Squares |
Mean Square |
F-statistic |
| Regression |
3 |
228.492 |
76.164 |
28.360 |
| Residuals |
168 |
451.176 |
2.686 |
|
| Total |
171 |
679.667 |
|
*****************
PSPP
*****************
Currently (Nov 2009) PSPP has
problems with getting output to a file. You can't cut and paste
or save the output anywhere. It is supposed to save to a file
called psppire.txt but I can't find that file anywhere on my
computer. I did a print screen, saved it to paint, and then saved as
jpg, and then inserted to this page. Awkward.
Return
to top
Click here
to return to the free software page
last updated 11/6/09
last verified 3/23/09
try packages using data sets here
http://pages.stern.nyu.edu/~jsimonof/classes/1305/pdf/excelreg.pdf