Upload
others
View
8
Download
0
Embed Size (px)
Citation preview
DOCUMENT RESUME
ED 201 298 IR 009 171
AUTHOR Ling, Robert F.TITLE Guidelines to the Use of Computers in Statistical
Instruction.SPONS AGENCY National Science Foundation, Washington, D.C.PUB DATE 78CONTRACT NSF-SED-76-12191NOTE 46p.
EDRS PRICE MF01/PCO2 Plus Postage.DESCRIPTORS *Computer Oriented Programs: *Computer Programs: Data
Analysis; Evaluation Methods: Higher Educatior:Instruction; *Mathematics Education; Questionnaires:Statistical Analysis: *Statistics
IDENTIFIERS *Statistical Packages
ABSTRACTThe introduction to this selective bibliography on
the use of computers in teaching statistics provides a brief reviewof the role of the computer in statistics: the role of statisticalpackages; how statistical packages should be used in instruction:science, statistics, and data analysis: and choice and evaluation ofstatistical packages. Items listed in the bibliography are concernedboth with instructional applications of computers and with theevaluation and comparison of statistical software. A copy of aquestionnaire seeking information on available statistical softwareis appended, together with selected responses. These questionnaireswere used to compile the data reported in the Index of PubliclyAvailable Statistical Software, by Kohm, Ryan, and Velleman (19771.(LLS)
************************************ ,********************************** Reproductions supplied by EDF a-e the best that can be made *
* from the or :Lai document. *
********************************* vt*********************************
ok.
U 5 0Epas4TWENT OF HEALTH.EDUCATION 6 WELFARENATIONAL INSTITUTE OP
DUCATION
r DOCUMENT HAS BEEN REPPD-_.:-D EXAC-_Y AS RECEIVED FR:W
-L ERSON .7P ORGANIZATION °RIG N.NG IT PO TS OF VIEW OR OPINIC NS"Ir.,D DO 7 NECESSARILY REP '4E-r OFFiC, NATIONAL INSTITUTE- OF:AT ION 'HON OR POLICY
UIDE___2,ES TO THE USE OF C _.IPUTEaS
L TTATISTICAL INSTRUC:ION
Robe= F. Ling
ABSTRACT. __L- ant aut.ho7.7 p. -.ri:_es -. .___, _
with an out= a _nf..n7...:Lon, ai: c. Li L-1:-; -d
references --,7 =for-7:A ..., on "LoH_:s dL
ted to the 1_ ...-J2s iLatis':- _ :___ ins: -_1L.-.
Robert F. Ling is ssor, Deparr erut -f Mathematical Sciences,
Clemson Universi':-7. Clemson, SC 29E:1- 7-reparation of this docu-
ment was supported 2 cart by a great 17.77a the National Science
Foundation (SED )1, titled "IlLa-=:..ic and Instructional
Services for Undelate Students 1.7: .:_atistics", directed by
Jerry A. Warren cf ta ,University Hampshire, Durham, NH.
2
rT I G
The use f the cc72- tcr Eta 7 =7...ica1 analyses has.aes= growing leaps ...A :.oands in t. cast decade. SomeLn:Lrect evid;:. 3 of tn a;Li grcwt a=e: the increased.=:..crtion of reperting compu:e= based results in
:77 ist ica 1 ' :nals; ta,l. large n...-tber c publicl y
suailable St _stical :.7:,ftware and parages (r ported by
Ryan, u ellema: (1117), e,q.) any f ich became7 a ilab le on.' within ti.e. five year: the c: -ttion of a
Lie:-.._ion on Statistica: cm in th :icanStatistical n ; ..!,LtiattL.:. of c-7-oups,
Ne-,:sletters, oe...-1:-Lal Ts:Fet: cs, e withpa rtic ula s7.:a 7 2 5 s..c!I as E AS an as
well as incr ET-S 'f ouLises,
tutorials, an': .2= i.j.7e tzr. paznlag..ss; and Dn.
The use of an: statist...cal pac,-:- 7es for
statistical _nstr-Icti..r.- -7 _rte ct her hand , has been rowiaqat a much slowe7 co.:par:son usaqe,whether measured .2 Sys :a andregular use of = 27- _s farfrom being univer:., and; lay s=:.11 be r-r- %sidere _ :e theexception rather t r. e, -L all of
instruction. Tne7c-3 Lam rea.zcns f 7 he slowgrowth in coputr-- izteJ. iTS..tzuctLon Twoof the maia ones ar:-. (1) tarry instruct:rs of s=27- .stics,who are knowledgea)le in ±rt siject anc its tra___::.ional
methods of instruct:L:7_, lre T.mfam _list , or ili-a-easein, the us of the .=ter jai eneral r the pa=7...Larsin the use of statL.st-I=1 rges1 cr effecti7e uss ofsuch facilities to Et. '.71e '...nstrun of statiscaltopics. consegaentI-, tEit material a.::d lecture nctes :hatare independent cf COMr titer and re...7-::-2Tt advances iL itssoftware are till ur. (2) if those instructorswho are knowledges'ble i statistic. comouting, =he
following type cf :.::.(eT'imet is often expressed, "it is
sufficiently difficu:.. t o :cie: the s=.1-..:_sticLi materia: in
a course without ha7.1-vc tc teach th!a students compl-ing(such statements are ott-ti ac7:-11:anief ty laments about thestudents' lack of tackc-ound interest in mathema:Lcs,statistics, or computimcv : -uld be impossible to t-iachthem the use of statisticc. _ cies (or the computer) andstill have time left to teac: t1%-.n statistics:"
Both the above reasGI: can be attributed to the samecause, to a large extent. 7."rlis cause might be termed"unfriendly computerese". II is quite apparent that
difficult-to-learn or unr7,==a1 languages associated withoperating systems and s::.-.ical packages have beendeterrents to statistical !:_amp._ting for many Instructors andtudents ilho otherwise wcuJ made. considerably more use
3
of them. Fortunately, the situation is steadily improving:package developers are generally paying more attention tothe userinterface aspects of their products; user manualsare better written; and there are recently developedsystems such as IDA and Minitab which are particularlysuitable for in conjunction with statistical instructionrecause of the simplicity of the larguages.
In this article, the author provides the reader with anoutline of basic information, as well as sources andreferences for further information, on topics directlyrelated to the use of computers for statistical instruction.Cie believe such information will be useful to those who arecontemplating the introduction of statistical packages as anaid to their teaching. Some, of the information should beuseful even to those whc are already using the computer intheir statistical instruction because they may not be awareof much of the existing rescurses and literature referencesassembled here.
2. THE ROLE OF THE CCMPIITER IN STATISTICS
The computer is playing a major role today in variousaspects of research, application, and instruct ion in
statistics. In view of the tretaendous power of existingcomputers and the availability of high quality software,statisticians not only call perform traditional computingtasks (associated with well established methodology) with
utmost ease accompanied by a high degree of numericalprecision, but can also apprr!ach old problems from new
points of view much of which wculd not have been possiblewithout the computer to perform the otherwise extremelytedious computatior.,:- song the major areas of statisticalapplications in wh,, .::omputer is an indispensible toolare:
A.--Monte Ca es and Simulation Experiments,
B.--Routine Statistical Analysis of Large Data sets,
C.--New Methodologies Involving Large Amounts ofComputing, e.g., Robust Methods, Iterative Methods,and Subsampling Methods,
D.--Exploratory Data Analysis and Model Building,
E.--Validation of Statistical Assumptions via ResidualsAnalyses and Graphical Methods.
4
3. THE POLE CF STATISTICAL PACKAGES
Current usage of statistical packages, whether measuldby the number of people engaged in those activities oramount of computer time used prc:ably falls within t
following categories. listed in decsing order of usage:
A.Descriptive and Tabular Ana Isis. Tools for Use-sin their preparation of of 17.eports invol7ig tabu_and graphical displays cf data or summariesdata.
B.--Tools' for statisticians to facilitateperformance of tasks in 2.AE.
C.--Aids to teachers in the instruction of statistics,
4. HOW STATISTICAL PACKAGES SHOULD BE USED IN INSTPUCTICN
We first propose some general guidelines on the use ofstatistical packages in instruction, independent of how theyare used to convey specific statistical concepts or to
illustrate specific methodologies. They are as follows:
A.Packages should siaply be used as a tool in theproper analysis of statistical data. A properanalysis must not to ccnstrained by what a
particular package can or cannot do. Often ananalysis may require the use of several differentpackages because of the limitations of capabilitiesin the "standard" packages. Sometimes, none of theexisting packages will deliver ;That needs to bedone on a particular problem, in which case specialprograms or routines must be written to perform thenecessary tasks.
B.--The roles of statistics, statistical packages, andtheir interface should be clearly established. It
is the duzy of the instructor to convey tc thestudents the 2scper ccncelts of statistics and the.prole r application of statistical methodology,regardless of what statistical packages areavailable, what procedures are available within aparticular package, cr what sorts of outputs areavailable within a particular procedure of a
particular package.
Statistical packages (or the instructica of
2e packages hemselves) shc-lld play.312 -cp or to the le ( and application
3 t _ Far tcc :.ften, -_ude:-. ts anizu t:.- e get too in vol7ed th the
n cf various sot twa :acka.gesla :ay acquire the skill,: tc Dt1 cf
to run usihg a:: ticalac _le they lack the skills tc
:ror.erly or to now theTitra::e7- should follow in an g
Or
The ; =.:.-epts of stat ics car 7
a: simple numeriial example=stay .c is pEcimages. Such __emonstrattons"pr:q-e"f -_ang but they =invariably 1E-
s t e _ tact better grasp of the th.F_ccetcon a :s have seen one or more 1-1',17rie,--_ caldem cr7 :1-s of those concepts. The funda7 tal
u ,--±,_1:i_ying a sampling' distributi;:h a_
conf-i =n :terval, the central limit theore..-c,fa aT: difficult theorems can ease anccnv".n--1:....-_ -7 be illustrated by numerical eT:-..-ample.T
to -.fical packages.
It is vet: di _._.cult, if nct impossible, to se t ::orth aset of guid&-iines Dn how to teach specific to: '-es ofstatisti=s a irq t , computer. To teach any panticulartopic, the .anal material which is optimal for anelemer ry eL cou_Tse will nct be suitable f or a:- audiencein at ii-..termc,:ria te ad vanced courser Hence, e7. en if thesame 1.-..-.7.7aatist :al pa. "-;:age is used tc aid the instr uction ofthe mach care must be given to how thecomps ion i rest:its are tc be presented or used bydiffer t olu_sses of students (at various levels Theredoes - par to be any sound, genera 1 rules, nor anycone ated source cf good examples. In subs ec=ion A ofthe arence section, a number of recent art:- des arecite: TF :ise articles deal with the actual uL..=': of thecompLier-i or statistical packages at various Levels ofinst-n-v =on, on a large number of topics. The list ofartic...l- is only intended to serve as a re pre. sentative
recent work on computer-based methods or -cater ialin stic al ins tructicn. Moreover, sire= manysta 1 packaces are being used in instructic onlythose art_cl es about "the use cf package X in teacb._---7-1" arecited. ;:rticles about "package X" or about st atiisticalpackages but without an emphasis on their use forinstru- tic are purposely omitted in order to keep t :le listmoderately small. While the ccmputer and stat.: sticalpackages a_.:e used to facilitate computational tasksassociation with the execution of statistical methodology,one important aspect of statistics -- its uncirlyingphilosophy -- and the impact of the computer on the
6
realization of t he phi: 7ophy, is of te: overlooked ii
instruction, especially :a elementary sta7 Lstical coursesWe shall brief17 tiscuss -"lat philosophy nc approach in th-7:next section.
EC STA r.C3, AND DAri.. ANALYSIS
Tracl textbookL: o: statistics and conventionalmethods c _ 'motion oft 1 short off the real goals 3::
statistic of prof c _lying and analysis o f r ea_
data. problems E =ten presen-i-ed in the form ofclear-cu: LL-defined E- = yes in confirmatory analysisthat bea: resemblano., = the problems one is likel.:to e nco the Feal .d.
St :al analysis is a delicate blend cf art andscience statistilta: data is seldom, if ever, a
one-pas Lf7-. In tc routine examination anddisplay pf Ls data for terical Lccuracy, edit andtransfo- -ecessary, s.tatisticia r. usually has to
carry analysis th h several iterations beforearr iv ir, ore tentative .:nclusicns. The use of the
compute struction cire- facilitates the teaching ofthe rea and proper a cach tc sta tistics. Such aphilosc approach i articulated in the followingexcerpt:
a aalvsis , and the parts of statistics whichadhere o it, must then take on the characteristics of ascience rather than those cf mathematics, specifically:
(1) Data analysis must seek for scope andusef loess rather than security.
(2) Data analysis must be willing to errmoderately of ten in order t hat inadequateevidence shall more often suggest the rightanswer.
(3) Data analysis must use mathematical argumentand mathematical results as bases f or
judgment rather than as basis for proof orstamps or validity." (p. 6)
"If data analysis is to be well done, much of it mustbe a matter of judgment, and theory' , whether statisticalor non-statistical, will have tc guide, not command ." (p. 10)
"The most important maxim for data analysis to heed,and one which many statisticians seem to have ve shunned, is
this: Far better an approximate answer to t he rightquestion, which is often vague, than an exact answer to the
7
wrong question, which can always be made precise. 13)
JOHN TUKET (1963) "The suture of Data Analys___:G" inAnnals of Matheratical Statistics.
What Tukey said fifteen years ago rerains todaybecause exploratory data analysis entails consid= morecomputation, through trials and refinements, thy__ its one-pass confirmatory analysis counterpart. Such an a: Droach ispossible, but impractical, without the Tcrt ofappropriate computing software. Eox(1976) rem .Ja-_ed thesame theme in a recent article in which he said
"In the inferential stage, the analyst amts as a
sponsor of the model. Conditional on the assum.:tf or of itstruth he selects the best statistical procedure for analysisof the data. Having completed the analysis, acwever, hemust switch his role from sponsor to critic. C.2nditionalnow on the contrary assurption that the r del. may beseriously faulty in one or more suspected or unsuspectedways he applies appropriate diagnostic checks involvingvarious kinds of residual analysis." (p. 793)
"The symptoms of (cookbookery) are a ten,-Lency to forceall problems into the molds of one or two routinetechniques, insufficient thought being given to the realobjectives of the investigation cr tc the relevance of theassumptions implied by the imposed methods." (p.
"Mathematistry is characterized by develoienttheory for theory's sake, which since it seldom toucles dr,with practice, has a tendency tc redefine the proble ratherthan solve it." (p. 797)
GEORGE EOX (1976) "Science and Statistics" inJournal of the American Statistical Association.
The iterative process of being a sponsor and a criticof various tentative models is most effectively carried outwith the aid of appropriate interactive statistical packagescr languages. The reason is that although the same steps ofanalysis can be carried out usirg packages run in the batch-mode, the long waits in turnaround between job submissionand completion generally inhabit (at least deter) the
statistician or the student from trying out a large numberof small changes in the analysis or exploring as manyalternative nodes of analysis as they would under an
inter active nviroment.
The interactive system IDA was created at a time and
8
8
enviroment when 6 hours turnaround for batch jobs wasnot uncommon, is that exploratory analyses could becarried out in c _ ::oticn with a course in data analysiswithout the fru---- =-..:_ons of long delays assoLiated withbatch computing. __mos its creation in 1972, it has notonly been used and effectively as an aid to
statistical but has also been a useful tool inmany projects of serj.ous research and real problem solving.Ling ang Roberta (1975) describes the interface betweeninteractive comp__-ing and data analysis as follows:
"I,,..:teractive computing articulates extremely well withthe requirements of enlightened statistical analysis, in
which the a nalwst examines data to help formulate anappropriate statfstical model, applies diagnostic checks tocriticize the mzrael, revises the model as necessary, andcontinues the process until he is satisfied with that he hasdone about as as he can within the given constraints oftime and space. Only then does the final statisticalanalysis ensue, and that is relatively cut and dried." (p.
411)
"Unlike many textbook descriptions of statisticalanalysis, and also unlike many statistical packages (even intimesharing), IDA does nct presume that an analysis need bedone in a fixed sequence of operations. In order to rAploitthe capatabilities of user-machine iteration, IDA breaksdown statistical operations into relatively small modulesthat can be executed in any sequence that is logicallypossible. This permits the user to take advantage of whathe has learned at each stage before deciding what to do
next. Except for certain obvious restrictions in the orderof execution of commands (for example, the user must firstenter his data before he can edit, display, or operate onit) , the user has complete freedom in going from any onecommand to any other. Such freedom is particularlyimportant and usef ul in exploratory data analysis for
regression, where, in the process cf deciding on a modelthat does justice to what seems to be happening in the data,the user generally needs tc examine residuals and fittedvalues for tentative models, try transformations, possiblydelete observations to obtain insight intc the effect of
outliers, and examine various plots anl numerical model-
adequacy checks, not in any specific order but in an orderthat appears appropriate for the problem at hand, dependingon the feedback and interpretation of intermediate results.Such a mode of analysis, though possible under other
systems, is generally much more cumbersome to execute thanin IDA." (p. 415)
ROBERT LING AND HARRY ROBERTS (1975) "IDA: An Approachto Interactive Data Analysis in Teaching and Research" in
Journal of Eusiness.
9
6. CHOICE AND EVALUATION CF STATISTICAL PACKAGES
A useful and up-tc-date reference cn the publiclyavailable statistical software packages is the ILdex ofStatistical Software, edited by Fobert F. Kohm, Thomas A.R/an, Jr., and Paul F. Velleman; available in microficheform in the Proceedings of the Statistical ComputingSection, American Statistical Associaticn (1977) . Itconsists of a compilation cf the responses by 56 packagedevelopers regarding the capability, availability,portability, cost, and other relevant information about eachpackage. In addition, abstracts cf the packages (written bytheir developers) are provided.
Kolm, Ryan, and Velleman describe the purpose andcontent of the Index as follows:
"The purpose of the index is to provide a singlereference capable of answering many cf the first questionsthat may arise when an individual is trying to determinewhich computer program might best serve his needs. To servethis goal, the index consists primarily of three parts.Part 1 is a listing of the general capabilities of theindexed programs. The goal of this portion of the index isto cross reference programs and capabilities, e.g., whichprograms do ANOVA, Siiple Data Descriptions, FactorAnalysis, etc. The second part of the index consists cfmiscellaneous details about each program. This portion ofthe index is designed to aid the reader in determiningwhether the programs of interest tc a particular user willrun cn his machine, in what computer language it is written,in what form is the program available, etc. The thirdportion of the index consists of the names and addresses cfthe developer, distributor cf the program, distributor ofthe documentation, and the person who completed the .
questionnaire reply. It alsc contains an abstract of thecomputer program written by the person who completed thequestionnaire, typically the developer. Virtually no
guidelines were given to the developer in the writing of theabstract, so they are quite different in appearance. Theabstracts provide the developers with a form in which theymay describe special features cf their program that are notadequately covered by the questionnaire. Many developershave chosen to describe their strcng points and thosefeatures which they feel make their programs unique."
The author believes the Index will serve its goals welland that the Index (and its future editions) will prove tobe an indispensible scurce cf information about theavailable statistical software. The questionnaire whichgenerated the Index is reproduced in Appendix A, and thetabulation of the responses is reproduced in Appendix H.
0
10
The Index makes no attempt to evaluate any of theindexed packages. There are many recent articles which dealwith the evaluation and comparison of packages or particularroutines within packages. A list cf such articles is givenin subsection B of the Reference section.
In choosing one cr more statistical packages forinstructional purposes, care should be given to the "ease ofuse", "handiness", or "friendliness" of the packages. A
package which has a complicated control language or datastructure will necessitate students spending much timelearning to use the package itself, thus leaving much lesstime learning about statistics. Existing packages leavemuch to be desired in their user-interface implementations."Unfriendly" systems and packages still abound, and theyshould be avoided as instructional aid in statistics,especially in courses taught at an elementary level. Manyuniversities find it necessary tc offer special 1-credithour courses just to teach students the control language ofone statistical package, such as SAS (which is one of themore "friendly" packages). The fear and frustrationstudents may experience in using less "friendly" packagescan be extrapolated from that.
7. OTHER SOURCES OF INFORMATION
Apart from journal articles, meetings, and otheractivities affiliated with scientific societies, there arenumerous special publications and special organizations thatprovide services and information that ale pertinent to theuse of the computer in statistical instruction. Some ofthese are described below:
A. Information cn how computers are used forinstructional purposes (not limited to statisticalinstruction) by 57 educational institutions ranging inlevel from elementary schools to major universities has beencompiled by the Human Resources Research Organization,published as the Academic Computing Directory (1978): TheDirector, not only identifies the "exemplary" institutions,,but provides information cn the computers they have, how
they are used, and the reasons the institutions wereselected as "exemplar" of academic computing. In addition,the name and address of a ccntact person at each institution- one who will answer questions from inquirers - is providedfor each entry. Copies of the Directory are available for$3.95 each frcm:
Human Resources Research Organization300 N. WashingTon StreetAlexandria, VA 22314
11
B. EDUNET is a national network of colleges anduniversities formed to promote the sharing cf computer-basedresources in higher education. The operation of EDUNET isoverseen by the Planning Council ct Computing in Educationand Research, which was formed in 1974. The PlanningCouncil is one of several special activities of EDUCOM - anon-profit organization established to further cooperativeefforts among institutions of higher learning. EDUNET makespossible the connection of any computer terminal to manycomputer centers across the country. Among the statisticalprograms and packages accessible under EDUNET (January,1978) are: APL, BNDP, EATAEDIT, INSL, MANOVA, MINITAB II,SPSS, TSP (at Stanford); BMD, DATATEXT, LIDA, TPL, TSP (atYale) ; FCST2***, FCST3***, FCST4***, STATNCV***,STATPROB***, STATTS * ** (at Dartmouth) ; MINITAB, STATJOB (atWisconsin) ; SCSS (at Notre Dame) ; and many other packages atother universities such as Cornell, NIT, Princeton, andRice. For information about available resources andservices, documentation, charges and accounting, and otherEDUNET matters, one may call the toll -free EDUNET Hotline(800) 257-9505, or write to:
EDUNET CentralP. 0. Boy 364Princeton, NJ 03540
and current information items about EDUNET aremembers (and tc others upon request) by the
ons EDUNET News and ECUCCM Bulletin.
C. The Symposium on the Interface of Computer Scienceand Statistics, held annually since 1967, consists of aseries of workshops cn topics involving statisticalcomputing. Papers presented at the workshops appear in theProceedings of the Annual Sumpcsium. Articles on the use ofcomputers in statistical instruction has appeared regularlyin recent Proceedings.
D. The Association for Educational Data Systems (AEDS)
is an organization which sponsors numerous computer relatedactivities in education on a variety of topics, including,computer-assisted instruction, computer-managed instruction,computer-assisted guidance, computer-assisted testing, and
so on. Although most cf its sponsored activities are notaddressed specifically to statistical instruction, many ofthem do apply tc computer assisted instruction in
statistics. Its affiliated publications (and annualsubscription rate) are: the AEDS Journal ($20) , the AEDSMonitor ($12) , and the AEDS Bulletin ($5) , all publishedquarterly. AEDS membership dues are $25 per year (student$10?. which include subscriptions tc the publications listedabove.
12
8. REFERENCES
A. Commuters in Statistical Instruction
Back, H.L. and Riley, A. (1975) , "A Note on the Use of aComputer In Generating Examples for Practical Classesin Biometry Teaching to Ncn-Mathema ticians," Bulletinin Applied Statistics, 2, 17-18.
Baker, f.B. and Rubner, V.P. (1973), Statistical ConceptsPackage, Department cf Educational Psychology,University of Wisconsin.
Bogyo, T.B. (1973), "The Use cf MANOVA in Teaching aGraduate Level Experimental Eesign Course," Proceedingsof Computer Science. and Statistics: 7th AnnualSym.posium on the Interface, Statistical Laboratory,Iowa State University, 252-253.
Bohrer, R. (1973) , "Computer Packages in TeachingStatistics: a SOUPAC User's View," Proceedings ofComputer Science and Statistics: 7 th Annual Sym_posiumon the Interface, Statistical Laboratory, Iowa StateUniversity, 274-278.
Bradley, Drake R. (1978) , "A n Interactive Data-Generatingand Answer-Correcting System for Problems inStatistics," Behavior Research Methods andInstrumentation, 10, 218-227.
Bradley, D.R., Hotchkiss, C.M., Duniais, S.T., and Shea, S.L.(1976,, "Computer Assisted Instruction in the SmallCollege," Proceedings of the .7th Conference onComputing in the Undergraduate Currirula, Department ofComputer Science, Texas Christian University, 205-213.
Brown, William R., Cock, Ida, and Unkovic, Charles M.,(1977), "Effectively Teaching Undergraduate SocialResearch via Computer- Data Analysis," Proceedings ofthe 8th Conference on Comaiters in the UndeIgraduateCurricula, Department of Computer Science, TexasChristian University, 323-300.
Butler, Michael D., Dwass, Meyer, Joiner, Brian L., andSwanson, James M. (1976) , "Report on the Roles ofComputers," in Modular Instruction in Sta ti Elle s,Washington,D.C.: American Statistical Association,14-19.
Cox, D.F. (1973), "The Use of the Statistical AnalysisSystem in Teaching," Proceedins of Computer Scienceand Statistics: 7th Annual Symposium on the Interface,
13
13
Statistical Laboratory, Iowa State University, 236-237.Craig, Frances B. and Eddy, William F. (1977), "Interactive
Data Analysis Through a Computer Network," Proceedingsof the Statistical Computing Section, AmericanStatistical Association, 169-173.
Dutiouchel, W.H. (1973), "Using MIDAS to Teach ElementaryStatistics," Proceedings of Comnu ter Science andStatistics: 7th Annual Symposium on the Interface,Statistical Laboratory, Iowa State University, 262-269.
Dunn, Robert M. and Gentlemen, Jane F. (1977), "InteractiveStatistical Graphics as Used in Data Analysis andTeaching," Proceedings of the Statistical ComputingSection, American Statistical Assoc ia tion, 202-205.
Eskin, G. and Montgomery, D. (1976), Data Analysis: Cases inComputer and Model Assisted Marketing. Teaching Notes.Stanford Barn, Palo Alto: Scientific Press.
Evans, D.A. (1973), "The Influence of Computers on theTeaching of Statistics," Journal of the RoyalStatistical Society, (A)136, 153-190.
Foster, F.G. and Smith, T.M.F. (1969), "The Computer as anAid in Teaching Statistics," ALplied Statistician, 18,264-270.
Frane, J.W. (1973), "Educational Aspects of the EMD7 and ENDSeries of Statistical Computer Programs: Point of Viewof a Developer," Proceedings of Comput?r Science andStatistics; 7th Annual Symposium on the Interface,Statistical Laboratory, Iowa State University, 238-242.
Gallant, A.R. (1973), "Some Arguments Against the Use ofStatistical Packages in Teaching Statistical Methods,"Proceedings of Computer Science and Statistics; 7thAnnual Symposium on the Interface, St atisticalLaboratory, Iowa State University, 223-225.
Geeslin, William E. (1977), "The Computer as an Aid forModular Instruction in Statistics," Proceedings of the8th Conference on Computers in the UndergraduateCurricula, Department of Computer Science, TexasChristian University, 347-350.
Gentleman, Jane F. ( 1976) , "Interactive Graphics in aTerminal-Equipped Classroom," Communications inStatistics, Part A - Theory and Methods, 1, 949-968.
Gentleman, Jane F. (1977) , "It's All a Plot (UsingInteractive Computer Graphics in Teaching Statistics) ,"the American Statistician, 31, 166-175.
14
14
Hartle y, H.O. ( 1976) , "The Impact of Computers onStatistics ," in on the History of Statistics andProbability, New York,N.Y.: Marcel Dekker, 419-442.
Haugh, L.D. (1973), "A Computer's Role in IntroductoryStatistics Courses," Proceedings of Computer Scienceand Statistics: 7th Annual Symposium cn the Interface,Statistical Laboratory, Iowa State University, 291-296.
Holmes, P. (1975) , "Using Computers in the Teaching ofStatistics," Ma thematical Gazette, 59, 228-245.
Hopper, M.L. and Cohen, S. (1976), "Statistics, Speakeasyand the Computer," Proceedings cf Computer Science andStatistics: 9th Annual Symposium on the Interface, LosAngeles, Calif.: Health Sciences Computing Facility,UCLA, 250-255.
Joiner, B.L. and Campbell, C. (1975) , "Some InterestingExamples for Teaching Statistics," Mathematics Teacher,68, 364-369.
Jowett, D. (1973) , "OMNITAE--A Tool for TeachingStatistics," Proceedings cf Computer Science andStatistics: 7th Annual Syniscsium cn the Interface,Statistical Laboratory, Iowa State University, 228-232.
Kossack, C.F. (1973) , "On the Use of the Computer in theTeaching of Statistics," Proceedings of Computerscience and Statistics: 7th Annual Symposium on theInter face, Statistical Laboratory, Iowa StateUniversity, 226-227.
Krutchkoff, R.G. (1973), "Why Teach Stochastic Simulation?"Proceedings of Computer Science and Statistics: 7thAnnual Symposium on the Interface, St atisticalLaboratory, Iowa State University, 210-212.
Lehman, Richard S., Starr, E. James, and Young, Kenneth' C.(1975) , "Computer Aids in Teaching Sta tistics andMethodology," Behavior Research Methods andInstrumentation, 7, 93-102.
Ling, R.F. and Roberts, H.V. (1975) , "IDA: An Approach toInteractive Data Analysis," Journal of Business, 48,411,451.
Maghsoodloo, S., and Bool, J.N. (1976), "On ResponseSurface Me thodology And Its Computer-AssistedTeaching," the American Statistician, 30, 140-144.
Main, D.A. (1971), "A Computer Simulation Approach forTeaching Experimental Design," in Proceedings of the79th Annual Convention of the American Psychological
15
Association, Washi ngton, D .C. American PsychologicalAssociation.
Margolin, Barry H. (1976), "Design and Analysis of FactorialExperiments Via Interactive Computing in APL,"'Technometrics, 18, 135-150.
Mead, R. and Stearn, R.D. (1973), "The Use of the Computerin the Teaching of Statistics," Journal of the RoyalStatistical Society , (A) 136, 191 -2C4. Discussions,205-225.
Meyer, Cecil H. (1977), "Use of a Computerized ForecastingModel in the Teaching of Basic Business Statistics,"Proceedings of the 8th Conference on Computers in theUndergraduate Curricula, Department of ComputerScience, Texas Christian University, 25-32.
Millward, R., Mazzucchelli, L., Magoon, S., and Moore, R.(1978), "Intelligent Computer-Assisted Instruction,"Behavior Research Methods and Instrumentation, 10,213-217.
Muller, M.E. (1970), "Computers as an Instrument for DataAnalysis," lechnometrics, 12, 259-293.
Quade, D. (1971) , "On Using a Conversional Mode Computer inan Intermediate Statistical Analysis Course," Review ofInternatioNal Statistical Institute, 39, 343-345.
Roberts, Harry V. (1978), "Statisticians Can Matter," theAmerican Statistician, 32, 45-51.
Rowe, K.E. (1975) , "SIPS as a Fart cf Statistical Computingin Teaching and Data Analysis," Proceedings cf ComputerScience and Statistics: 8th Annual Symposium on theInterface, Health Sciences Computing Facility, UCLA,56-60.
Ryan, T.A. Jr., Joiner, and Ryan, B.F. (1975),"Teaching Statistics witb Minitab II," Proceedings ofthe 6th Conference on Computers in the UndergraduateCurricula, Department of Computer Science, TexasChristian University, 195-204.
Ryan, T.A. Jr., Joiner, B.L. , and Ryan, B. F. (1976) , MINITABStudent Handbook, N. Scituate: Duxbury Press.
Bywick, Thomas (1975) , "Increasing Student Interest by theUse of Interactive Computer Simulation," BehaviorResearch Methods and Instrumentation, 7, 103-104.
Scalzo, Frank, and Hughes, Fowland (1977), "IntegratingPrepackaged Computer Programs Into An Undergraduate
16
Introductory Statistics Course," Proceedings of the 8thConference on Computers in the Undergraduate Curricula.Department of Com=:ater Science, Texas ChristianUniversity, 331-338.
Schatzoff, M. (1968), "Aplicaticns cf Time-shared Computersin a 3'i.-Atistics Curriculum," Journal of the AmericanStatistical Association, 63, 192-208.
Searle, S.R. (1973) , "How Little Computing Need We Teach toStatistics Majors?" Proceedings of Computer Science andStatistics: 7th Annual Symposium on the Interface,Statistical Laboratory, Iowa State University, 204-209.
Simon, J.L., Atkinson, D.T., and Shevokas, Carolyn (1976),"Probability and Statistics: Experimental Results of aRadically Different Teaching Method," AmericanMathematical Monthly, 83, 733-739.
Smith, D. J. (1977) , "Teaching Basic Statistics by Means ofSimulations," Bulletin in Applied Statistics, 4,1,
88-94.
Sterrett, A. and Karian, Z.A. (157E) , "A Laboratory for anElementary Statistics Course," the AmericanMathematical Monthly, 85, 113-116.
Storlie, J. (1978), "Microcomputers in Public Education:Bccn or Bane?," AEPS Monitor, 17, 10,
Swanson, James M. (1976), "An Interactive StatisticsLaboratory," in Mcdular 'Instruction in Statistics,Washington,D.C.: American Statistical Association,90-96.
Tanis, Eliot A. (1977) , "A Computer -Based Laboratory forMathematical Statistics and Probability," Proceedingsof the 8th Conference on Computers in the UndergraduateCurricula, Department of Computer Science, TexasChristian University, 334-346.
Thurmond, J.B. and Cramer, A.O. (1972), "Toward the OptimalUse of Computer Simulations in Teaching ScientificResearch Strated," Proceedings of the Conference on
Computers in the Undergraduate Curricula, Atlanta,Georgia: Georgia Institute of Technology.
Tracy, Robert J. (1976), "A Computer-Aided Procedure fcrProducing Interesting Assignmerts fcr Students of
Applied Statistics," Behavior Research Methods andInstrumentation, 8, 413.
Wallace, D.L. (1969), "Computers in the Teaching of
Statistics: Where ar the Main Effects?," in
17
Statistical Computation, R.C. Milton and J.A. Nelder,eds., New York: Academic Press, 3149-361.
Warren, R.D. (1973), "Using SPSS in Statistical MethodsCourses," Proceedings cf Computer Science andStatistics: 7th Annual Symposium on the Interface,Statistical Laboratory, Iowa State University, 2113-2114.
Zahn, D.A. (1978), "The Statistics Panel of UNAP,"Proceedings of the Statistical Ccmputing Section,American Statistical Association, 1978.
B. Evaluation and Comparison of Statistical Software
Allen, I.E. and Velleman, P.F. (1917) , "The Handiness ofPackage Regression Ecutines," Preceding§ of theStatistical Computing Secticn, American StatisticalAssociation, 95-101.
Allerback, K.R. (1973) , "Data Analysis Systems: a User'sPoint of View," Social Science Information, 10, 23-35.
Avery, K.R. and Avery, C.A. (1975) , "Design and Developmentof an Interactive Statistical System (SIPS) ,"Proceedings of Com_p_gter Science and Statistics: 8thAnnual Symposium on the Interface, Health SciencesComputing Facility, UCLA, 49-55.
Anderson, R.E. and Coover,E.R. (1972), "Wrapping Up thePackage: Critical Thoughts on Applications Softwarefor Social Data Analysis," Compiters and theHumanities, 7, E1-92.
Berk, Ronald A. (1977) , "Survey cf Integrated StatisticalComputer Packages," Behavior Research Methods andInstrumentation, 9, 277-280.
Eryce, G. R., Francis, I., and Heiberger, R. M. (1976),"Statistical Software Evaluation in the USA," COMPSTAT,Second Annual SEmposius cn Computational Statistics,327-334.
Carmer, S. G. (1975), "One Statistician' s View of ConsumerEvaluation of Statistical Software, " Proceedings cfComputer Sciences and Statistics: 8th Annual Symposiumon the Interface, Health Sciences Computing Facility,UCLA, 149-154.
Driscoll, Eileen, and Francis, Ivor (1977), "Some Measuresof Regression Package Perfcrmarce," Proceeding§ of theStatistical Computing Secticn, American Statistical
18
18
Association, 190-195.
Forsythe, A. and Hill, M. (1975) , "Design cf Experiments forComparative E "aluaticn cf Statistical Packages,"Proceedings of the Statistical Computing Section,American Statistical Association, 17-20.
Francis, I. (1973), "A Comparison of Several Analysis ofVariance Programs," Journal of the American StatisticalAssociation, 68, 860-865.
Francis, I. (1975), "The Ncvice With a Statistical Package:Performance Without Ccmpetence," Proceedings ofComputer Science and Statistics: 8th Annual Symposiumon the Interface, Health Sciences Ccmputing Facility,UCLA, 110-114.
Francis, I. and Heiberger, R.M. (1975), "The Evaluation ofStatistical Program Packages- -The Beginning,"Proceedings of Computer Sciences and Statistics: 8th
Annual Symposium on the Interface, Health SciencesComputing Facility, UCLA, 106-109.
Francis, I., Heiberger, R.M., and Velleman, Paul F. (1975),"Criteria and Considertions in the Evaluation ofStatistical Program Packages," the AmericanStatistician, 25, 52-56.
Gentle, J.E. (1975), "Cemparisons cf Statistical Packages byUsers Having Some Familiarity With Computing andStatistics," Proceedings of the Statistical ComputingSection, American statistical Association, 1975,114-117.
Heiberger, R.M. (1973), "Statistical Computing ThroughStatistical Packages: An Introductory Course,"Proceedings of Computer Science and Statistics: 7thAnnual Symposium on the Interface, StatisticalLaboratory, Iowa State University, 218-222.
Heiberger, R.M. (1975a), "A Procedure for the Review ofStatistical Packages and Its Application to the UserInterface With Regression Programs," Proceedings ofComputer Sciences and Statistics: 8th Annual Symposiumon the Interface, Health Sciences Computing Facility,UCLA, 115-121.
Heiberger, R.M. (1975b), "Activities and Plans of theCommittee on Evaluation cf Statistical ProgramPackages," Proceedings of the Statistical ComputingSection, American Statistical Association, 1975, 1-2.
Hill, Maryann (1976a), "Prepared Discussion of the Workshopon °Evaluation of Statistical Program Packages',"
19
Pr ceedings of Comsuter Science and Statistics: 9thAr amposium on the Interface, Lcs Angeles, Calif, :
.ith Sciences Computing Facility, UCLA, 149-1'
(1976b) , "Where Next in Software Eval.Proceedings of the Statistical ComputingAmerican Statistical Association, 176-177.
Hoaglin, D.C. and Andrews, D.F. (1975) , "The Rev. ofComputation Based Results in Statistics," The ,canStatistician, 29, 122-126.
Kohm, R.F. (1975), "Preliminary Report: Index of AvailableStatistical Software," Proceedings of the StatisticalComputing Section, American Statistical Association,1975, 3-9.
Miller, John E. and Francis, Ivor (1976), "A ComputingNovice With a Statistical Program Package: InteractiveVersus Batch," Proceedings of Computer Science andStatistics: 9th Annual Symposium on the Interface, LosAngeles,Calif.: Health Sciences Computing Facility,
,A, 231-236.
Polla,,, Leonard I.. and Schnittjer, Carl J. (1977) ,
Principal-axis Factor Analysis: a Comparison of FourSelected Computer Program Packages ," Education andPsychological Measurement, 37, 207-212.
Ryan, T.A., Jr., Rohm, B.F., and Ryan, B.F. (1975),"Interactive Statistics," Proceedings of ComputerSciences and Statistics: 8th Annual Symposium on theInterfaced Health Sciences Computing Facility, UCLA,66-70.
Schucany, Minton, P.D., and Shannon, B.S.,Jr. (1972),"A Survey of Statistical Packages," Computing Surveys,4, 65-79.
Swanson, J.M. and Riederer, S.A. (1975 , "IMP and SHFIMP:Small Interactive Mimics of OMNIT AB," Proceedings of.
Computer Sciences and Statistics: 8th Annual Symposiumon the Interface, Health Sciences Computing Facility,UCLA, 84-84.
misted, Ronald A. (1976), "User rocumentation and ControlLanguage: Evaluation and Comparison of StatisticalComputer Packages," Proceedings of the St atisticalComputing Section, American Statistical Association,24-30.
Velieman, P.F. (1975), "Interactive Computing forExploratory Data Analysis 1: Display Algorithms,"Proceedings of the Statistical Computing Section,
Z.)
American Statistical Association, 142-146.
20
Velleman, P.F. and Welsch, R.E. (1975) , "Some EaluationCriteria for Interactive Statistical Program Puckages,"Proceedings of the Statistical Ccrlputing Section,American Statistical Asscciaticn, 10--2.
Velleman, P. F. and Welsch, R. E. (197n), "On EvaluatingInteractive Statistical Program Packages,"Communications in Statistics, E, 5, 17-206.
Velleman, Paul F. and Allen, I. Elaine (1976), "ThePerformance of Package Regression Routines UnderStress: a Preliminary Trial of A Regression EvaluationMethod," American Statistical Aszcciation, 1975Proceedings of Statistical Ccuuting Section, 297-304.
Velleman, P.F., Seaman, J. , and Allen, I.E. (1977) ,"Evaluating Package Regression Routines," Proceedingsof the Statistical Computing Section, AmericanStatistical Asscciation. 82-92.
Velleman, P.F., Seaman,J., and Allen, I.E. (1978),"Evaluating Accuracy and Cost cf Computer Programs forMultiple Regression: Theory, Methods, and Applicationsfor Widely Distriuted Statistical Packages."Unpublished manuscript.
Wilkinscn, L. and Dallah, G.E. (1977), "Accuracy of SampleMoments Calculations Amcrq Widely Used StatisticalPrograms," American Statistician, 31, 128-131.
APPENDIX A
Available Statistical Software Questionnaire
reproduced (with permission) from
Kohm, Robert F., Ryan, Thomas A. Jr., andVelleman, Paul F. (1977), "Index of PubliclyAvailable Statistical Software." microfichein Proceedings of the Statistical ComputingSection, American Statistical Association.
22
AVAILABLE STATISTICAL'SOFTWARE QUESTIONNAIRE
1-1 Complete name for the package or program?
1.2 Short name or abbreviation for the program?
1.3 Principal developer:NameOrganization_DepartmentAddress
Telephone (
1.4 Distributor if different from (1.3) above:NameOrganizationDepartmentAddress
Telephone (
1.5 Distributor of documentation if different from (1.4 ):
NameOrganizationDepartmentAddress
Telephone (
1.6 Person responsible for preparation of this reply if different from (1.3).
NameOrganizationDepartmentStreetCity State Zip
Telephone (
1.7 Date of last program release and identification (e.g. Version 5.3)
Date Identification
1.8 Expected date of next release.Date
Page 2
2.1 Doea L'ilx program run in batch mode?
(a) Yes(b) No
2.2 If your prcgram is interactive, can it run as a (One or more answers)
(a) Fully Interactive program with prompting questions.(b) Fully interactive program with answers immediately
available, but without prompting questions.(c) As a control card checker (i.e.
essentially interactive checking of controlcards for large batch-like analyses).
2.3 Program availability. (One or Lore replica)
Not available for export to other computation centers?(b) Passively distributed to other centers?(c) Actively distributed to other centers?(d) Available through time-sharing system?
2.4 If your program 1s available for export, how is it available?
(One or more answers)(a) Under a rental agreement. Coat
(b) For purchase. Coat
(c) Through some computer association such as SHARE,VIM, CUBE, DECUS, CONP!3IT, etc.
2.5 If your program is available for export, how is it transported?
(a) In a form which is compatible with one computer brand only.
(b) In some readily transportable language (e.g. ANSI Fortran).(c) In separate versions for different computer brands.
2.6 If your program is available for export, approximatelyhow many centers have received copies of your program?
2.7 Which of the following documentations does your program have?
(One or more answers.)
(a) Primer intended for novice users or students withworked out examples.
(b) Reference manual.(c) On line "HELP" selectively available °on request."(d) Implementation or systems programmers guide.(e) Other
2.8 Please check all compilers or interpreters required.
(a) Fortran II(b) Fortran IV
ANSI(c) Not verified(d) Verified on Bell Labs PFORT verifier(e) Verified on cther verifier
(f) Extended FORTRAN IV(g) COBOL(h) PL/1(i) Assembler(3) BASIC(k) APL(1) Algol 60(m) Algol 68(n) Other (please specify)(o) Special compiler (please specify)
2.9 On which computer makes has the program been
successfully run.
Page 3
Manufacturer Model Operating system (if important)
2.10 Is your program:
(a) Under continuing development.(b) Being maintained only.(c) No longer being maintained.
2.11 Are local program modifications permitted?
(a) Yes(b) No
2.12 Are local additions of capabilities permitted?
(a) Yes(b) Ho
2.13 Has your program or package (or a technique unique toyour program or package) ever been reviewed or described in a widelydistributed journal or at a society meeting?
(a) Yes (please list these on a separate sheet of paper)
(b) No
Page 4
3.1 Rank your program's strengths with respect to the following
applications (1 for strongest, 2 for second strongest etc.).
Do not rank inappropriate items.
(a) The analysis of data.(b) Data tdanipulation, editing, table building, file handling.
(c) Monte Carlo simulations(d) Teaching(e) Other
3.2 Which of the following would best describe your program?
(Check only one.}
(a) A stand alone main program.(b) A collection of separate main programs.(c) A special purpose statistical package or set of
programs (e.g. a linear models package).(d) An integrated general purpose system of statistical
routines running under a monitor.(e) A set of subroutines or similar groupings of code.
3.3 What is the principal intended audience of your program
in terms of statistical experience? (check only one)
(a) Complete novices with little or no statistical background.
(b) Naive users with one introductory survey course in
statistics.(c) A user with moderate exposure to statistics coming
from advanced courses in statistics or from periodic
exposure to statistical methodology.(d) An advanced or sophisticated user possessing an
extensive background in statistical method and theory.
3.4 How would you describe your principal intended audience
in terms of computer experience? (check only one)
(a) Novices with no previous exposure to data processing
or computers.(b) Naive tiaera who may have used the computer somewhat
(in a course for example).(c) Users with moderate exposure to data processing
techniques and familiarity with either some languages
or with experience in using statistical packages.
(d) Advanced or sophisticated utters with programmingexperience and/or data processing expertise.
3.5 For what general fields is your package intended?
(One or more answers)
(a) Biological sciences(b) The social sciences(c) The engineering and physical sciences
(d) Business and economics(e) General statistics(f) Other(s)
2
Page 5For the Questions 4.1 - 4.30 please circle the most appropriate answer using the
following replies:
C Capability
L Limited
D Documented
F Feasible
I Insufficientor "Blank"
P Planned
0 Obsolete
The program or package has sufficientcapabilities in this area to be consideredWI a feature.
The program or package has some capabilitiesin this area, but they should be consideredas limited.
The.featurn can be easily accomplishedusing the documentation supplied with theprogram, but is not a standard ( "built -in ")option.
The feature can be accomplished by userswith some expertise in the area.
The program or package has insufficientcapabilities in this area to be classified above.
The program or package has either limitedor no capability in this area, but programchanges are planned to include this areain the "C6 classification within one year.
The program developer considers the programor this aspect of the program to be obsolete.
Programs may be listed in several categories. (For example, anonparametric analysis of variance program would be listedboth in Analysis of Variance and in Nonparametric Statistics.)
Page 6THE GENERAL CAPABILITIES AREAS
pata_ Manakement
CLDFIPO 4.1 File building and manipulation.
CLDFIPO 4.2 Variable and category labeling.
C L D F I P 0 4.3 Data transformations (standardization, logs,arithmetic, etc.)
CLDFIPO 4.4 Matrix computations and manipulations.
CLDFIPO 4.5 Sorting and matching.
Editing
C L D F I P O 4.6 Case selection/exclusion.
CLDFIPO 4.7 Consistency checking (simple and compound logical checks)
CLDFIPO 4.8 Automatic error correction (consistency correction,imputation, etc.)
Computations
CLDFIPO 4.9 Simple data descriptions and statistics (means,atd.dev, histograms, ttest, simple linear regression,one way ANOVA, bivariate plotting, etc.)
CLDFIPO 4.10 Multiple linear regression.
CLDFIPO 4.11 Analyxis of Variance (balanced data andtraditional designs).
CLDFIPO 4.12 Analysis of Covariance (balanced data andtraditional designs).
C L D F I P 0 4.13 Canonical correlation and principal component analysis.
CLDFIPO 4.14 Factor Analysis.
Page 7
C L D F I P 0 4.15 Multivariate analysis of variance anddiscriminant analysis.
C L D F I P 0 4.16 Non-linear regression.
C L D F I P 0 4.17 Cluster analysis.
C L D F I P 0 4.18 Simple analysis of multi-way tables (percentages, means,atd. dev., etc.)
C L D F I P 0 4.19 Loglinear analysis of multiway tables.
CLDFIPO 4.20 Other analysis of categorical data.
C L D F I P 0 4.21 Variance estimation for complex sample designs.
C L D F I P 0 4.22 Time series analyses (spectrum analysis,Box-Jenkins, forecasting, etc.)
C L D F I P 0 4.23 Data smoothing and curve fitting(exponential smoothing, splines, etc.)
C L D F I P 0 4.24 Bayesian statistics.
C L D F I P 0 4.25 Nonparametric statistics.
C L D F I P 0 4.26 .:Random number generation and simulation.
Output
C L D F I P 0 4.27 Table printing (multi-way tables: formatted,category labels, etc.)
CLDFIPO 4.28 Graphics (versatile displays of data andintermediate results on printer, teletype,or special graphics hardware. Note: histogramsand plots in which users have no control overthe choice of variables, scales, or plot symbols,are here defined to be Limited° L.
C L D F I P 0 4.29 Results of computations available Immediately(not in a subsequent run) as input to otherprocedures (e.g. residuals treated as data toother procedures.)
CLDFIPO 4.30 Results of computation and intermediate results(e.g. OR decompositions, correlation matrices, etc.)output in machine-readable form (disk/tape/cards)..
- - -- Thank you ----
APPENDIX B
Selected Questionnaire Replies
reproduced (with permission) from
Kohm, Robert F., Ryan, Thomas A. Jr., andVelleman, Paul F. (1977), "Index of Publicly
Available Statistical Software," microfichein Proceedin s of the Statistical Com utinSection, American Statistical Association.
9. Simple data description
10. Multiple linear
regression
11. Analysis of variance
12. Analysis of covariance
13. Canonical correlation
i principal component
14. Factor analysis
15. Multivariate ANOVA
A liscriminant analysis
16. Nonlinear regression
17. Cluster analysis
18. Simple analysis
of' multi-way tables
19. Loglinear analysis
of multi-way tables
20. Other analysis
of categorical data
21. Variance estimation
(complex sample designs)
22. Time series analysis
23. Data smoothing
4 curve fitting
24. Bayesion statistics
25. Nonparametric statistics
26. Random generation
It Simulation
GENERAL CAPABILITIES LISTINGPAGE 9
1. Pile manipulation
2. Variable labeling
1 1 1
I I I3. Transformations
1 1 1111., 4. Matrix manipulations
11 I 111 I I5. Sorting I matching
III 1111.IIIIIIIIII
IIIIIIIIIIIIIII 1 11
IIIIIIIII11I11IIIItlllllltll1III III I I IIIIII 1 27. Table printing
I I I I I I I I 1 I(multi-way tables)
1 t I I I I I I I I28, Graphics (versatile)
IIIIItIllIIItIII1IIItIIIIIl1111111III III 1I1111IIIIII1I1111111IIII111111111111111111111111I I I I I I I I I I I 1 I 111111 31. Run batch
I 1 I I I I I I I I I I I 1111111 32, Run interactively
111111111111111111111I I I ' I I I I I I I I 11111111 33. Statistical level
II I III III I I I 111111111 34. Computing level
1111 III 11 11111111111111111111111111111111111
6. Case selection/exclusion
1. Consistency checking
of data
8. Auto error correction
29. Reuse of computations
in same run
30, Storage of computations
Program
2 2 2 2 2 2 2 1 1 1 1 1 1 1 2 2 2 3 3 3 3 3
6543210987654321091234567878901234
MIT-SNAP
AFSTAT
AOD
The IMSL Library
TABSTAT
TSP /DATATRAN
GLIM
BHPMOPBMD
WRAPS
LINWOOD and NONLINWOOD
SAS
CS
P-STAT
Minitab II
SPSS
RUMMAGE
HPSTAT PACS
CADA Monitor
IMPRESS"
SCSS
NTSYS
MATCAL
C-TAB II
31
Abstract
pages I.D.
1 c icc11ci1 1ce b a b
1c to 1 1 0 1 ace cc lccll lac sac
1 I alcallcl f fcc lcoccdtloac lob
ccpca lacccpococcc lcall 1 1 1 co cc1 ccfcciccl c cc fnacc
coals, c 10 lcc ccc cc we a da1 cot occc lc' 1 lac bcb
cppe d co repo° ccC f poaccocococcoloc 'mecca a b
f 3 c o a c o c c o 11 cif 1 1 1 1 ab
1 1 lcollooc c oc c
c 1 1 c c c o f f l o o bb
f 1 f floffoccoocoocoo.cococcococc boa
1 c o c a 1 c I a ofolflaccoc000coccocnaobp1 c p o cooccacc0000co coca bbb
lc p p p Ilccplollo f pool boa
10 pppc pocll0000c lacfcl a ae
1 at 1 calf 1a t 11,a Dab
to coropIpproacoccalotelf ca.:Jonas b
ppcp p tl cconcppoc cocpnabacc dd cdodoffdddooccoddol ccdcnaaa
1 1 c cc000i el c line1 c 1 c o Illocc oleo ob
oocco c oc acob
18 - 20 1977.001
21 - 22 1977-002
23 - 25 1977-003
26 - 29 1977-004
30 - 31 1911-005
32 - 34 1917-006
35 - 35 1977-007
36 - 36 1971.008
37 - 41 1977-009
42 . 44 1977-010
45 - 46 1977-011 .
47 - 48 1917 -012
49 51 1971-013
52 - 55 1977-014
56 - 57 1977-015
58 - 59 1971-016
60 - 61 1971-011
62 - 63 1977-018
64 - 65 1977 -019
66 - 67 1977-020
68 - 68 1977.021
69 - 70 1971-022
71 - 71 1977-023
72 - 72 1977.024
73 - 73 1977-025
9. Simple data descriotion
10. Multiple linear-
regression
11. Analysis of variance
12. Analysis of covariance
13, Canonical correleion
& principal component
14. Factor analysis
15. Multivariate ANOVA
discriminant analysis
16, Nonlinear regression
17, Cluster analysis
18. Simple analysis-- -
of multi-way tables
19. Loglinear analysis
of multi-way tables
20. Other analysis
of categorical data
21. Variance estimation
(complex mole designs)
22, Time series analysis
23, Data smoothing
& curve fitting
24. Weston statistics
25. Nonparametric statistics
26, Random generation
Simulation
GENERAL CAPABILITIES LISTING'PAGE 10
II
II
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
I I
IIII
II
1 I
III I
I I
I I
I I
I I
I I
I I
I I
1. File manipulation
2. Variable labeling
3. Transformations
4. Matrix manipulations
5. Sorting i matching
6. Case selection/exclusion
7. Consistency checking
of data
8. Auto error correction
21. Table printing
(multi-way tables)
28. Graphics (versatile)
29. Reuse of computations
in same run
30. Storage of cohputatfona
31. Run batch
32. Run interactively
33. Statistical level
34. Computing level
Program
2 2 2 2 2 2 2 1 1 1 1 1 12 2 3 3 3 3 3 Abstract
6 5 4 3 2 I 0 8 6 5 4 3 2 0 9 1 2 3 4 5 6 7 7 8 9 0 1 2 3 4 pages I.D. I
EFAP, COMM, LISREL III
MULTIQUAL
MULTIVARIANCE
LOGOG
NORMOG
EXPAX
MOCA
TSAR
GENSTAT
SOUPAC
ROSEPACK
STATIIrao
LOGLIN
Speakeasy
OMNITAB II
AESTH
DATAPAC
DATAPLOT
PROBPAC
ALSTAT
OSIRIS
CENTS-AID
TSERIES
EXPLORII
STATJOB
3J
ctflfffdd t dc
1 1 tl f
1
dd ci ft
1 1 ccfff1 1 pp
cp
11 cplc
c 0 o a cc cb
c c cb
c ccccc cc a c cc cb
o c c eb
c c bb
c c c bb
c c cc bb
fottofcdecceccfloolcl ca
occpcfccecelcoodcdfccco cb
cllcocicceoloodcd lice cb
c c c o d c c c lc° acc
fff decocco ccc cc nabs
P 1 c1 c c c a c b
f f cc fc fecccpccdlc ccoc aca
pf ff f ffccclpcccdd peed benodcd cdddclddncbb
cPPPPPPPPecfrPec leer cc
f f t 1 p o c 11 a C c
nbeclIcellc cc col bba
ccdocicccccociccdclipc cc
c 1 cc c c be
cc c cll 1 c cb
c c 1 1 cb
c leftcocccc ccldccec bca
e D 11
f f f
I I
c
Pc
74 - 75 1977-026
76 - 76 1977-027
77 - 77 1977-028
78 - 78 1977-029
79 - 19 1977-030
80 - 80 1977-031
81 - 81 1977-032
82 - 82 1977-033
83 - 83 1977-034
84 85 1977-035
86 - 86 1977-036
87 - 87 1917-037
88 - 88 1977-038
89 - 90 1977-039
91 - 92 1977.040
93 - 93 1977.041
94 - 95 1911 -042
96 - 97 1977-043
98 - 99 1917 -044
100 *11 1911 -045
102 - 105 1977-046
106 - 106 1977-047
107 - 107 1977-048108 - 108 1977-049
109 - 110 1977-050
34
9. Simple data description
10. Multiple linear
:egression
11. Analysis of variance
12. Analysis of covariance
13. Canonical correlation
& principal component
14. Factor analysis
15. Nultivariate ANOVA
& discriminant analysis
16. Nonlinear regression
17. Cluster analysis
18. Simple analysis
of multi-way tables
19. Loglinear analysis
of multi-way tables
20. Other analysis
of categorical data
21. Variance extiestion
(complex sample designs)
22. Time series analysis
23. Data smoothing
& curve fitting
24. Bayesian statistics
25. Nonparametric statistics
26. Random generation
& Simulation
GENERAL CAPABILITIES LISTINGPAGE 11
I I
IIIIIIIIIIIIIIllllllllllllll
I 11
File manipulation
2. Variable labeling
3. Tran$forsations
4. Matrix manipulation:
5. Sorting 1 matching
6. Case selection/exclusion
7. Consistency checking
of data
8. Auto error correction
! II I
27. Table printing
I ! I I l l t l l I I I I I I 1 I I(multi-way tables)
I I I I I 1 1 1 1 1 I I I I I I I I I28. Graphics (versatile)
IIII111111111111I I I 1
I I 1 11
29. Reuse of computations
1111 I 111 I I I I I I I I I I t Iin same run
I I I I II I II I II I I I I I I I I I I I30. Storage of computations
1 1 11 II I! 1 1 1 1
11 I I I 11 I 11 I 11 I I I I I I I I I I I !31. Run batch
I 11111111 I I I I I I I I I I 1111 32. Run interactively
I !II I 11 I 1 11 1 1 I 1 1 I
1111 I I I t I I I! I 11 111111 33. Statistical level
11111 I I I 11 I III 11 I I I I I I I I 1111111.-- 34. Computing level
111111111111111111111111111111111 111111111111111111111111111111I
Program
TPL
ITABS
GRAFSTAT
COCENTS
IDA
Data-Text
2 2 2 2 2 2 2 1 1 1 1 1 1 2 2 2 3 3 3 3 3
654321098765 432109123:567878901234
cc
c fcc
Abstract
pages I.D. A
c
cf d coo
C C C
C C C cc c ccp pccccffcdflecfnaba
1 c cd cb
c c of oleo cc lc cocoa ecblc c c cloaca cc oc Ionic as
111 - 111 1977.051
112 - 113 1977.052
114 - 114 1977.053
115 - 115 1977-054
116 - 117 1977-055
118 - 119 1977-056
30
Selected Questionnaire Replier, Page' 12
IDENTIFICATION 71- 71- 17- 11- 11- 77- 17- 11- 17- 11- 71- 77- 71- 71- 71- 77- 71- 11- 71- 77-
1 ? 3 A 5 6 1 8 9 10 11 12 13 14 15 16 11 18 19 20
2.1 RUN BATCH
T E S O R K O I nI yl yI nI yI yI nI nI nI !lint nI nlyI nI nInI nI yI yI
2.2 INTERACTIVEIIIIIIIIIIIIIIIIIIIIIA FULL/PROMPTS I laIaI IaIaI I I I I I I IaI I I I 111181
B FULL /HOPROHPTS I b I I I I I b I b I I I I I I b I b I b I b I I I I I
C CONTROL CHECKINZIIIiIIcIIIIIcIIIIIIIcIII2.3 AVAILABILITT I I I I I I I I I I I I I I I I I I I I I
A NO EXPORT I I I I I I I I I I I I I I I I I I I I I
B PASSIVE DISTRIBIDIIIIIIIIIIbIIIIIIIIIIC ACTIVE DISTRIBI IcIcicI 1010'0101cl IcIcIcIcIcIoIcIcicID VIA TIMESHARINGI IdIdI IdIdI I I I I Idl IdI I IdI I II
2.4 HOW AVAILABLEIIIIIIIIIIIIII.IIIIIIIA RENTAL I laIaIaI III IaI I I I IaIraI IalaI I IIB ONE TIME COSTIbI IbI I I IbI IbIbI I I I I I I I IbIbIC COMPUTERASSOCIIIIIIIIIIIIcIIIIIIIII
2.5 TRANSPORT HETHODI I I I I I I I I I I I I I I I I I I I I
A ONE BRAND COMPUTIal IaI I I I I I I IaI IaIal I I I IaIIB PORTABLE LANGUAGI I I I I IbI I b I b I b I IbI I I IbI IbI IIC DIFFERENT'VERSIOI I c I I c I I I c I 1 o I c I I c I I I c I I c I I I c I
2.6 NUMBER OF SITESIIIIIIIIIIIIIIIIIIIIIA 0 -10 lalalaI IaI Ill' I I IaI I IaI I I I IaII
11-99 II I I III IbIIII I I IbIIIblIbIC 100 OR MORE I I I 101 101 I IcIcI IcIc1 I IcIe1 I II
2.7 DOCUMENTATION I I I I I I I I I I I I I I I I I I I I I
A PRIMER I I IaI IaIIIIIaIa1 I IaI IaIaIaIII lollREFERENCE IbIbIbIbI IbI IbIbIbIbIbl bIbIbIbIbIbIbIbI
C O N L I N E H E L P I I IcI IcIc1c1 IoIoI I I Id' I I I I IID IMPLEM7N GUIDE I I I I I I I IdIdIdI I IdIdI IdIdldl Idl
E OTHER IIIIIIIIIIIIeIIeIIIIIII2.8 COMPILER & INTERI I I I I I I I I I I I I I I I I I I I I
A FORTRAN II I I I I I I I I I I I I I I I I I I I I
B FORTRAN IV IbIbI IbI IbTbIbIbIbI IbI IbIbIbI IbI IIANSI 111111111111111111111C NOT VERIFIEDIIIIIIclIcIIIIIIIIIIIIID PFORT IIIIIIIIIIIIIIIIIIdIIIE OTHER VERIFIERIIIIeIIIelIIIII IeIIe11 IIIF EXTENDED FORTIVIIflIfIIIIIIIIIIIIIIIIIG COBOL IIIIIIIIIIIgIIIIIIIIIIH PL/1 1h1IIIIIIIIIIIIblIIIIIII ASSEMBLER IIIIIIIIIIiIIIIIIIIIIIJ BASIC 1111111111111111111.11j'
APL IIIIIkIIIIIIIIIIIIIIIIL ALGOL 60 I I I I I I I I I I I I I I I I I I I I I
M ALGOL 68 I I I I I I I I I I I I I I I I I I I I I
11 OTHER IIIIIIIIIIIIIIIIIIIIIIJ SPECIAL COMPILERI I I I I I I I I I I I I I I I I I I I I
Selected Queetionneire RepliesPage 13
IDENTIFICATION 77- 77- 77- 71- 77- 17- 77- 77- 77- 11- 77- 77- 77- 77- 11- 71- 71- T7- 77- 77-
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 11 18 19 20
2.10 DEVELOPMENT I 1 I 1 I I I I I I I I III
A CONTINUING I aI aI al aItlaIaIaIa1 I aI aI aI eI aI aI aI aI aIaI
B MAINTAINED ONLY I I I I IIIIIIbIIIIII I I I I I
C 00 LONGER MIXT I I I I I I I I I I I I I I I I I I I I I
2.11 LOCAL MODIFICATI I I I I I I I I I I I I I I I
TES OR NO InInIn I ylyIyIyIyIYIYInIyIyIyInIy I yIyIyIyI
2.12 LOCAL ADDITIONSI I I I I I I I I I I I I I I I I I I 1 1
TES OR NO InIn IyIyIyIy In IyIy IyInIyIyIy I y IyIyIyIyIyI
2.13 REVIEWED 111111111111111111111YES OR NO InInIy IyIyIyIy IyIy1yIyIyIy Iyly IyIyIiIyIy1
3.1 RANK STRENGTHSIIIIII1 11111111111111A DATA ANALYSIS 11'1 I 1I1I 2121112111113111111121211111112 IB MANIPULATION I. I 2 I 2 I 4 I 1 I 3 1 3 I 1 I 3 I 3 I 2 I I 2 I 2 I 1 I 4 I 2 1 3 I 2 I 3 I
C SIMULATION I I 4 I 1 2 I I 4 I I 141 II II I 1 3 1 4 I IIID TEACHING 1 2 1 3 1 3 1 3 1 3 1 5 1 2 1 1 2 1 2 1 I 2 I I I 3 I 1 1 3 I 2 I 11 IE OTHER 111111111111111 1 3 1 I I I I I I
3.2 BEST PROGDESCIIIIIIIIIIIIIIIIIIIIIA MAIN PROGRANIIIIIIIaIIIIIIIIIIIIIIB SEP HAIN PROGRAMIIIIIIIIII0IIbIIIIIIIblIC SPECIAL PURPOSEIIIIIIIIcIIIIIIIIIIcII:DINT. SYSTEM IdIdIdI IdIdI I IdI I I IdIdIdIdIdI I IdI
E SUBROUTINES 1IIIeIIIIIIIIIIIIIIIII3.3 STATISTICAL EIPIIIIIIIIIIIIIIIIIIIII
A NOVICE IeI I I I I I I IaIaI I I I I IaIaI IaIIB NAIVE I I I I I I I I IbIbl IbI I IbI I I I IbI
C MODERATE I IcIcIelcI IcIcIcIcI I Ic1c1 I I lei IID ADVANCED IIIIIIdIIIdIdIIIIIIIII1 1
3.k COMPUTER EXPER, I I I 1 I I I I I I II
A NOVICE I I I I I IaI I I I I I 1st I !alai I IaI
8 NAIVE 1bI IbI I I lb1 IbIbI lb! IbIbI I IbIbIIC MODERATE I Icl 10101 I IcIcIcIel I I I 1 I I I IID ADVANCED IIIIIIIIIdIdIIIIIIIIIII
3.5 GENERAL FIELDSIIIIIIIIIIIIIIIITIIIIA BIOLOGICAL SCIENI I I IaI I IsIa1a1a1 IlIal I III 1411I1I8 SOCIAL SCIENCES I I IbIbI I IbIbIbI.bIbIbIbIbIbIbIbIbIbIbIC ENG A PHYSICAL! I I IcI I IcIcI I I IcIo1 I IcI IcIoIID BUSINESS A ECONI I IdIdIdIdI IdI I IdIdIdIdIdId1 I IdIIE GEN. STATISTICS1 e I e I e I e I e I I eI eIeI tIeIeI eIeI eIeIe1 eIeleIF OTHER IfIIIIfII I I I I I I IfIIIIIItI
Selected Questionnaire RepliesPage 414
IDENTIFICATION 71- 77- 71- 77- 11- 77- 71- 77- 77- 77- 77- 77- 77- 71- 77- 77- 71- 11- 77- 71-
21 22 23 24 25 26 27 28 29 3D 31 32 33 34 35 36 31 38 39 40
2.1 RUN BATCH 111111 I I I I I I I I I I I I I
YES OR NO I yl yInI nInInInI nInInInI n1 ni nInI yIyI yii I nI
2.2 INTERACTIVE I I I I I I I I I I I I I I I I I I I
A FULL/PROMPTS laIaI I I I I I 1 I I I I I I IalaIaIallB FULL/NOPROMPTSIblIIIIIIIIIIIIIIIIIbbIC CONTROL CHECKINOI I I I I I I I I I I I I I I I I I I
2.3 AVAILABILITIIIII 1 I I I I I I I I I I I I I I
A NO EXPORT IIIIIIIIIIIIIIIIIIIB PASSIVE DISTRIB I D I IbI 11111111 Ibl IbI I IbI
C ACTIVE DISTRIBI IcI IcIcIcIcicInIcIcIcI IcI lei I InIclD VIA TIMESHARINGIdIdI 111111111 Idl I I Idldldll
2.4 HOW AVAILABLE I I I I I I I I I I I I I I I I I I I I I
A RENTAL
B O N E T I M E C O S T lb! IbIbIbI IbIbIbIbIbIbIbI IbIbI IbI IbI
C COMPUTERASSOCIIIIIIIIIIIIIIIIIIIII2.5 TRANSPORT METNODI I I I I I I I I I I I I I I I I I I I I
A ONE BRAND COMPUTIlIaIaI I I I I I I I I Ial I'll I I 1 IIB PORTABLE LANGUAGI I I I I I I I II IIblIl Ibl Ibl Ibl
C DIFFERENT VERSIOI I I IcIcIcIcI IcIcicI I I c I lc] I I c I I
2.6 NUMBER OFSITESIIIIIIIIIIIIIIIIIIIIIA 0 -10 I I 1 Ial I I I I IaIalal I I Ialal I IIB11 -99 IbIbI I IbIbIbl lb1 I I IbIbIbI I Ibl IIC 100 OR MORE II IdII II I c I I I I I I III I Inlel
2.7 DOCUMENTATION I I I I I I I I I I I I I I I I I I I I I
A PRIMER Isla! I I I I III I I I label I IaIaIaIalB REFERENCE lb! bIbI blblblbl bIbIbl biblbl blbI I blblbIblCON LINE HELP IcIcI IcIcIcIcI IcI I I I I I IcI IcIal1D IMPLEMEN GUIDEIdIdIdl I I I I I I I I IdIdIdI I IdIdldlE OTHER I I I I I I I I 1 I I I I IeI Ielel leIel
2.8 COMPILER & INTERI I I I I I I I I I I I I I I I I I I I I
A FORTRAN II I I I I I I I I I I I I I I I I I I I I I
B FORTRAN IV I IbIbi I I I I I I IbIbIbl IbIbl I I IbI
ANSI I I I I I I I I I I I I I I I I I I I I I
C NOT
D PFORT IIIIIIIIIIIIIIIIdIIIIIE OTHRR VERIFIERIIIIIIIIIIIIIIIIIIIII
F EXTENDED FORT IVI I I IfIfIfIfIfIfIfI I I I I Ifl I I IIG COBOL I I I I I I I I I I I I I I I I I I I I I
H PL/1 I I I I I I I I I I I I I I I I I I I I I
I ASSEMBLER IIIIIIIIIIIIIIilliIIIIIIJ BASIC IjIIIIIIIIIIIIIIIIIIIIK APL I I I I I I I I I I I I I I I I I I I I I
L ALGOL 60 I I I I I I I I I 1 I I I I I I I I I I I
M ALGOL 68 I I I I I I I I I 1 I I I I I I I I I I I
MOTHER IIIIIII1IIIIIIIIIInIII0 SPECIAL COMPILER1IIIIIIIIIIIIIIIIPIIII
IDENTIFICATION
Selected Questionnaire Replies Page 15
77- 11- 17- 71- 17- 77- 17- 77- 11- 77- 71- 71- 11- 11- 17-
21 22 23 24 25 26 21 28 29 30 31 32 33 34 35
71- 71- 17- 71- 71-
36 37 36 39 40
2.10 DEVELOPMENT I I I I IIIIII1 I I I I I
A CONTINUING IaIdIeleI IaIaIdIaIaIal I laIsIa I a' laid'B MAINTAINED ONLYIIII I bI II 1 I IbIbI II I I I 1
C NO LONGER MAIKTIIIIIII I I I I I I I I I I I I I
2.11 LOCAL MODIFICATI I I I I I I I I I I I I I I I I I I 1 I
YES OR NOIyIyIyIyIyInTylylyIyIyIyIyInlyly I n lyInIy1
2.12 LOCAL ADDITIONSI I I I I I I I I I I I I I I I
YES OR 0 I y I y I y I y I y 1 n IyIy Iyly IyiyIyIyIyIy I y I y Iy111
2.13 REVIEWED I I I I I I I I I I I I I I I I I I I I I
YES OR NO Iy IyIy InlyIn Iy ly Iy In In In IyIyly IyIy ly ly Iy1
3.1 RANI STRENGTHSIIIIIIIIIIIIIIIIIIIIIA DATA ANALYSIS 1 2 I 3 1 1'11 1 1 I 1 I 1 I 1 111111 I 1 I 1 1 111 1 1 I 1 I 1 1 1 1 1 I
1 1 MANIPULATION 1 3 1 1 1 I 3 I I I I I I I I I 2 I 2 1 2 I 1 3 I 1 2 1 3 1
C SIMULATION 111111111111114111 5 I 114 ID TEACHING 11121 12121212121212121 1313141214121312 1E OTHER IIIIIIIIIIIIIII 3 I I 2 1 I I I
3.2 BEST PROGDESCIIIIIIIIIIIIIIIIIIIIIA MAIN PROGRAM I Is! 1 IaI IaIaIaIaIaIaI III I laIaI IIB SEP MAIN PROGRAMIlblIIIbIIIIIIIIIIIIIIIC SPECIAL PURPOSEIIII 1 1 I I I I & 1 1 1 1 lIcIIIIIDINT. SYSTEM IdIdIdl I I 1 I I I I I Idl Idl 1 I I IdlE SUBROUTINES IIIIeII1IIIIIIIIIIIIII3.3 STATISTICAL EIPIIIIIIIIIIIIII I I I I I I I
i NOVICE IIIIIaIIIIIIIIIIIIIIIIe1B NAIVE I I I 1 I I I I 1 IbIbIbl I I 1 Ibl I IbIC MODERATE I I IcI IclelcIoIcI I I IcicIoIcI IcIolcID ADVANCED IIIIIIIIIIIIIIIIIIIIdI3.4 COMPUTER BIM IIIIIIIIIIIIIIIII 1 I I I
A NOVICE III I I I I I I I I I I IalI I Ial lale1B NAIVE I I IbI IbIbIbIbIbIbIbIbl IblbI I Ibl IIC MODERATE IIIIcIIIIIIIII I .1IcIIIIID ADVANCED I I I I I I I I I I I I I I I I I I I I I
3 . 5 GENERAL F I E L D S I I I I I I I I I I I 1 1 1 1 IIIIIIA BIOLOGICAL SCIENI I I e I e I a I o I a I a I e I 14III 14Ialal 141111B SOCIAL SCIENCESIDI IbIbIbIbIbIbIbI Iblbl IbIbIbI IbIbIIC INGAPHYSICALI I I IcIoIcIoIcI I I I I IcI 1c1c1o1010ID BUSINESSISCONI I I IdIdIdIdIdIdl IdIdI I IdIdIdIdIdIIE GEN. STATISTICSI I I IeleIeIelel I I 1 IeIeleI IeI IelelF OTHER IIIIIIIIIrIfIIIIIIIIIIfI
44
Selected Questionnaire Replies
IDENTIFICATION 77- 77- 77- 77- 77- 77- 77- 77- 77- 77- 77- 77- 77- 77- 71- 77-
41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
2.1 RUN BATCH II:IIIIIIIIIIIIIIT E S O E N O I aI nI yl nI nI nI nl nI nI nI nI nI yI nI yI nI
2.2 INTERACTIVE I I I I I I I I I I I I I I I I I
A FULL/PROMPTSIIIIIII I I I I I IITai IaIIB FULL /NOPROMPTS I I I b I b I b I I I'l IbI I I I I IIC CONTROL CHECKINGIDIIIIIIIIIcIIIIIII2.3 AVAILABILITTIIIIIIIIIIIIIIIII
A NO EXPORT IIIIIIIIIIIIIIIIIPASSIVE DISTRIBIbIIIIIIIIIIIbIIIIIC ACTIVE DISTRIIII IcIcIcIcIcIcIcIoIcIoI IcleIcIcID VIA TIMESHARINGIII I I IdIdI I IdI I 1411 /I
2.4 HOW AVAILABLE I I I I I I I I I I I I I I I I I
A RENTAL IIIIIIIIIIIIIaIII a.I8 ONE TIME COST IbIbIbIbIbibIbIbI bIbIbIbIbIbIblIC COMPUTER ASSOCIIIIIIIIIIIIIcIIII2.5 TRANSPORT METHODII/iIIIIIIIIIIIIIA ONE BRAND COMPUTI I I I I I I a I I I a I a I I I I I a I
B PORTABLE LANGUAGI b I b I b I b I b I I I b 1 b I I I b I b I b I b I I
C DIFFERENT VERSIOI I I I I I c I I I IIIIIoIcII2.6 NUMBER OF SITESIIIIIIIIIIII
A 0 -10 I I I fi l a I a i I I a I I I 1.1 : I I IIB 11-99 I b I b I I I I IbI I b I b I I I I I b I b I
C 100 OR MORE IIIIIIcIIIIIolIloIII2.7 DOCUMENTATIONIIIIIIIIIIIIIIIII
A PRIMER I IaIalaI lei I I I Is! IaI I I # I
B REFERENCE I IbIbI IbIbIbIbIbIbIbIbI IbI IbIC ON LINE HELP I I IcIcicI I I I I I I lei lollD IMPLEMEN GUIDE I I I I I I d I d I I I d I I I I d I d I I
E OTHER InIeI I I lei I I IeI I I I leII2.8COMPILER&INTERIIIIIIIIIIIIIIIIIAFORTRANII I I I I I I I I I I I I I I I I I
B FORTRAN IV I I I I I b I b I I b I b I I I b I b I I b I b I
ANSI 11111111111111111C NOT VERIFIED I I I I I I I I c I c I I I I I I 43 I I
D PFORT I IdIdIdIdI I I I I I I I I I I I
EOTHERVERIFIERIIIIIIIIIIIIIIIIIFEXTENDEDFORTIVIIIIIIIIIIIIIIIIIG COBOL IIIIIIIRIIIIIIIgIIIli PL/1 I I I I I I I I I I I I I I I I I
I ASSEMBLER I I I I I i I i I I I I i I I I I I I i I
J BASIC IjIIIIIIIIIIIIIIIIK APL I I I I I I I I I I I I I I I I I
L ALGOL 60 I I I I I I I I I I I I I I I I I
II ALGOL 68 I I I I I I I I I I I I I I I I I
II OTHER III I I I I I I InInIInII II0 SPECIAL COMPILERI I I I I I I I I I I I I I I I I
e
Selected Questionnaire Replies
IDENTIFICATION 77- 17- 17- 71- 77- 77- 77- 77- 77- 77- 77- 77- 77- 77- 77- 77-
41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
10 DEVELOPMENTIIIIIIIIIIIIIIIIIi CONTINUING IsIaIaIstaIaIll I I 'aisle' IsIaIB MAINTAINED ONLYI I I I I I I IbIbIbI I I IbI IIC NO LONGER MAINTIIIIII1IIIIIIIIII11 LOCAL MODIFICATI I I I I I I I I I I I I I I I I
T E S O R N O IyIy IyIyIyIyIn /yIyIyiy IyIn ly lyIy1
12 LOCAL ADDITIONSI I I I I I I I I I I I I I I I I
YES OR NO IyIyIyIyIyIy In lyIyIyIy Iy IyIy IyIyI
,13 REVIEWED I I I I I I I I I I I I I I I I I
YES OR NO Iy Iy In InIy Iy Iy ly Ty Iy Iy Iy Iy ty Iy ly/
,1 RANK STRENGTHSIIIIIIIIIIIIIIIIIi DATA ANALYSIS 1 1 1 1 1 1 1 1 1 1 1 1 I I I 1 1 1 2 1 2 I 1 1 3 I 1 I 1
B MANIPULATION I 131 I I 12111 I 1211 1 112111412 IC SIMULATION I 1 2 1 IIIIIIIIII 3 I I 5 I 1
D TEACHING 12 I 4 I 2 I I I 1 1 2 1 2 1 I 1 1 4 1 2 I 3 I 3 I
E OTHER I I 1 3 1 1 1 I I I 1 1 I 3 I I I II 2 I I
.2 BEST FROG DESCIIIIIIIIIIIIIIIIIAMAIN PROGRAM tat I I I I I laIal IaI I I I IIB SEP MAIN PROGRAMIIIIIIIIIIIIIIbIIIC SPECIAL PURPOSEI I IoIcI I I I I I I laic' I IIDINT. SYSTEM I I I I Idldl I I Id! I I I IdIdIE SUBROUTINES 118111111.111111111.3 STATInICILEIPIIIIIIIIIIIIIIIIIA NOVICE IIIIIIIIIIIIIIIIslB NAIVE IbI I I IbI IbI I I I I IbI I IIC MODERATE I IoIcIoI lot IcloIcIoIcl IoIoIID ADVANCED I I I I I I I I I I I I I I I I I
.4 COMPUTER EXPER.IIIIIIIIIIIIIIIIIA NOVICE I 1 I 1 lei I I I taIaI IsI I leiB NAIVE 1 'DI I I I I I IbIbI I I I IbIbIlC MODERATE I IoI,oIcl toIoI I I 1 I0I I I IID ADVANCED I I I I I I I I I I I I I I I I I
5 GENERAL FIELDSIIIIII1 1
I BIOLOGICAL SCIENI a I I I I I IsIaIsIaIaI I s I I a I I
SOCIAL SCIENCESIDI I I I IbIbIbIbIbIbIbI IbIbIb1C ENG & PHYSICAL I o I I I 1(31 I IcIoIcI I I I I IID BUSINESS & ECONIdl I I /dIdI IdIdIdId/ I IdIdIIE GEN. STATISTIESIeIeIeIeI lel tot IeIeI I I Ie1 I
F OTHER I I I I I I I I I I I I I I I I I