HOME QUICK SEARCH AUTHORS KEYWORDS HELP BUGS

SUGI 1983 Papers - Index

178 papers. Source: SAS Users Group International conference proceedings.

Back to Conf
go 1984
go 1982
# Title
×		 Authors
×		 Abstract
×		 Keywords
×		 Extras Pages
1 A Bibliographic Citation System for CMS SAS John R. Gleason This system consists of a series of SAS macros designed to permit storage, retrieval, and man agement of bibliographic citation.:;, The macros are intended for relatively unsophisticated users of CMS/SAS. It is assumed that such users have access to an editor (e.g., XED!T) and a format ter (e.g., SCRIPT), but have relatively little working knowledge of SAS data library operations. One set of macros performs input of new citations fram rather free-form sequential files, with important features (e.g., date, author) of citations demarcated by user-modifiable signals. Citations to be input may (optionally) include keywords, and notes or connnentary of arbitrary length. Macros are available which allow the user to search through stored entries by date, author, keyword(s), words contained within the citation itself, and by portions or combinations of these elements dataset, regression analysis, proc step source 3
2 A Capacity-costing Methodology Michael Zambruski This paper describes a hybrid technical-financial procedure for allocating cost in a large data center. It is designed as a modeling tool to permit IIWHAT-IF" analyses for various configurations, levels of costing. etc. The use of modular programming, independent tables, and matrix output allows the modeler maximum flexibility in allocating cost according to changing technical and financial considerations. The first segment of the procedure focuses on a technique for determining primary system capacity. The second segment generates the cost-allocation mechanism, which assesses cost to users based on their individual impact on capacity cpu time, regression analysis source 10
3 A Comparison of Several Interactive CMS SAS Techniques Michael S. Lajiness It was decided that the best approach to the management of a simple interactive session involving novice computer users was to use a menu display format. The question that arose however was which of several techniques should one choose to facilitate the menu selection and subsequent program execution proc step, infile statement source 6
4 A Diagnostic Tool for Mixed Models with Application to Negative Estimates of Variance Components .. Ronald R. Hocking covariance, proc mixed, scatter plot, correlation, fixed effects, main effect source
5 A Dynamic Report Writer: Fitting One Structure to Multiple Outputs Jesse Gary,
Ann Kennedy Handler		 proc step, proc sort, proc transpose source 6
6 A Financial Plotting System for Analysts and Managers Laura S. Vendeland A major responsibility of financial analysts is to present the organization ' s financial picture to management in a concise and descriptive manner. At Texas Instruments, this responsibility has been translated into a Financial Plotting System for the collection, calculation, and graphing of monthly and year-end financial data. Designed for the non-programmer, the system functions in an in teractive environment and produces graphs according to user specifications. This paper out lines the development and structure of the system, its use by the financial analyst and business man ager, and the advantages of its operation missing value, proc step, dataset source 6
7 A Macro for Including a Response Grid on SAS/GRAPH G3D Plots John Hackman proc step, proc sort source 4
8 A Macro for Maximum Likelihood Fitting Log-Linear Models to Poisson and Multinomial Counts with Contrast Matrix Capability for Hypothesis Testing Gary G. Koch,
Maura E. Stokes		 The purpose of this paper ;s to describe a useful and readily implemented macro in PROC MATRIX for fitting log-linear models to small arrays of counts by maximum likelihood. Under lying distributions can be either Poisson or multinomial. Model parameter estimates are produced by applying the Newton-Raphson itera tive solution to a set of initial estimates based on weighted least squares. Goodness-of fit statistics are calculated, and a capacity exists for testing hypotheses concerning the model parameters. Examples are provided to illustrate a broad range of applications likelihood function, weighted least squares, poisson distribution, covariance, maximum likelihood estimation, proc step, confidence interval, multinomial distribution, odds ratio, estimate, proc freq source 6
9 A Multi-Evaluative Approach to the Analysis of Clinical Laboratory Data William A.T. Archambault Jr.,
James J. Tiede		 laboratory data, wilcoxon signed-rank, rank test, sample, proc reg, clinical trial, proc freq, kruskal-wallis test, parallel group, proc step, regression analysis source 6
10 A Multivariate Methodology for Exploratory Research Charles E. Rich factor analysis, proc step source 3
11 A Note on Using PROC NLIN for Function Optimization Marvin D. Troutt This note shows a simple device by which PROC NLIN may be used for quick and convenient ex ploratory function optimization. Although not as precise as exact codes and programs for non linear programming, many of these are less than user friendly. especially for casual model ex ploration. The proced~re is useful for overde tenmined systems of equations in particular. Some possible remedies for its theoretical shortcomings are mentioned proc step, proc nlin, regression analysis source 2
12 A Procedure for Estimating Parameters in Nonlinear Random Models Paulette Johnson,
George A. Milliken		 Consider an experiment where a nonlinear continuous function has been fitted to the data frcm several treatments or populations where the function depends upon a random variable-with un known mean and unknown covariance matrix. A two-step procedure. based on asymptotic theory, is developed to obtain estimates of the elements of the covariance matrix. The procedure is then applie& to a nonlinear one-way classification mocel and an example is presented sample, proc nlin, covariance, regression analysis source 4
13 A Program to Read SPSS-X' System Files Grant Blank,
Andrew A. Norton		 An extended version of SPSS*, named SPSS-X, will replace the existing SPSS release in early 1903. The new system files are substantially different from older SPSS files and they cannot be read by PROC CONVERT. Thus, conversion of the new files will require writing the file in raw data form, constructing SAS control CARDS, and reading the file into SAS with all the associated problems and cost of debugging, typing variable names, input formats and labels for hundreds or even thousands of variables. To avoid this arduous process, we have written a program, named SPSSXSAS, which will read the dictionary in the SPSS-X file and use that information to create. automatically, a $AS dataset. SPSSXSAS will convert variable names, labels, print formats arld missing values; and, in a valuable extension of the capabilities of PROC CONVERT, it will convert value labels and documents input statement, dataset, character variable, numeric variable, proc step source 5
14 A Proposal for Evaluating SAS Skills of Potential Employees Ann S. Allen,
Alvin M. Best III,
C. Royce Claytor		 in=, dataset source
15 A SAS News Facility Craig Dickstein,
J. Michael Shaughnessy		 sample, proc step, sas log source 4
16 A SAS Procedure for Accessing a. Model 204 Data Base Richard L. Gimarc A new procedure, PROe M204, is presented. This procedure allows a SAS user to build a SAS data set from data extracted from a Model 204 file. PRoe M204 couples the powerful data retrieval facilities of Model 204 with the data management, statistical, and graphical procedures of SAS. With this procedure, a SAS user can access data stored in a Model 204 database with little or no knowledge of Model 204. PROC 204 does not allow a SAS llser to access Model 204 data that he would otherwise be forbidden to access. defined through Model unauthorized access. The access controls 204 are used to limit A data dictionary facility is included as an integral part of PHOC M204. The dictionary is a SAS data set built and maintained by PROC M204 which maps Model 204 field names to SAS variable names for a particular Model 204 file or subfile. Performance data is presented which compares the various modes in which PROC M204 can be used proc step, dataset, cpu time, variable, data dictionary, if-then-else, proc print, sas log source 5
17 A SAS Program for the Extended Mantel-Haenszel Procedure Alan C. Fisher,
Daniel M. Wildman		 raw data, proc freq, clinical trial, proc step, dataset source 6
18 A SAS Program to Score the Michigan Tests of English Language Proficiency and Aural Comprehension Charles E. Rich macro definition, proc step source 6
19 A SAS Testscoring and Item Analysis Program Charles E. Rich do over, proc freq, proc step source 6
20 A SAS-based Marketing Information Service Debora J. Frye proc step, index source 5
21 A SAS-based Materials Tracking System K. Thomas Carr,
Roger B. Glaser,
Clark M. Roberts		 sas base, dataset source
22 A SAS-based VM/CMS Charge-Back System J.D. Hill At the Texas Instruments Lewisville Interactive Computer Center (LICC), a SAS-based accounting system has oeen developed to process accounting data for the purpose of billing in-house users. SAS was chosen to implement the accounting system for simplicity and ease of modification. The LICe IS VM/CMS operati n9 system automatically generates accounting records for many of the sys tem resources used. Additional accounting records are generated by special utility programs to allow billing for those resources not included in the basic accounting records. The SAS-based account ing system processes the records from the various sources, merging the data with additional informa tion such as division, cost center, and account number in order to produce a record for each re source used by each user. These records are then processed to produce journal entries to be sent to the Group Accounting System to bill the users dataset, cpu time source 6
23 A SAS-generated Backup Facility Barbara Belasco,
Suzanne Ives		 User error and system crashes necessitate periodic backups of SAS data sets. Although the corporate computer operations at Sandoz backs up clinical data sets weekly, we, as system userS, realized that to provide for data integrity we needed to backup data sets daily. Commercial backup software is costly and so we decided to use SAS to create the backups. The automated backup facility implemented at Sandoz uses SAS to read the system catalog and generate the backup job stream. The UNIX operating system is used to automatically send the job to the IBM mainframe at a predefined time and to store its output in a prenamed file libname statement, proc copy source 4
24 A Strategic Planning Control System Laura S. Vendeland Strategic planning, whether at a corporate or de partment level, requires a reliable control system to insure the accuracy. efficiency. and quality of its operations. At Texas Instruments, this need saw the deve 1 opment of an interact i ve. user-friendly Strategic Planning Control System (SpeS) to aid management in evaluating future fundable projects throughout the annual operations review cycle. The fol1owing paper describes the development and structure of the system, its use during the review cycle, and the advantages of its operations dataset, bar chart source 6
25 A User's View of SAS/IMS-DLI Emilio A. Icaza dataset, data step source
26 A User-friendly Disk Management System Thomas R. Hoffman,
Linda Tanis		 A disk management system has been developed that automatically backs up SAS data libraries and partitioned data sets to tape. These OS data sets are cataloged as members of generation data groups. Only SAS data libraries that have been modified since the previous backup are copied to tape. This feature not only provides the user distinct backups, but reduces both tape space and the number of entries on the system catalog. Partitioned data sets are not copied to tape unless they have been accessed since the previous backup. AlIOS data sets and their attributes are included in a SAS data library {DMSSASDLj. This library contains separate SAS data sets for SAS data libraries, partitioned data sets, and other OS sets. Inactive SAS data libraries and partitioned data sets are deleted from diSk by the disk management system, but maintained in the DMSSASDL libname statement, proc step, proc contents source 6
27 A User-maintainable SAS/RAMIS Interface for CMS Paul W. Fingerman dataset, proc step, data step, sysrc source
28 Accessing Vendor and User-written Subroutines with SAS Richard A. Boro,
Elaine S. Pretz		 character variable, proc step, proc print, character constant, binomial distribution source 6
29 Adverse Reaction Incidence Tables for Clinical Trials Donna R. Park,
Stephen J. Ruberg		 sample, proc step, crf, proc freq, by group processing, clinical trial, dataset source 5
30 An Algorithm for Producing Exact Maximum Likelihood Estimates for an ARIMA(p,q) Process Carol M. Reilly likelihood function, moving average, time series analysis, ma, covariance, maximum likelihood estimation source 3
31 An Analysis of Time-Series Data with Autocorrelated Errors .. Mary S. Hirsch,
E. Moye Rutledge		 PROC AUTO REG was used as the technique to analyze the data from an experimental septic tank filter field system. The \\'ater level within the system was seasonally dependent and the error terms serially correlated. The predicted lagged variable was included among the regressors after an instrumental regression used the other independent variables and their lagged values as predictors. The results from the autoregression analysis will be used to indicate the expected frequency of system failures from climatic stress predictor variable, proc freq source 3
32 An Automated Data Base to SAS/GRAPH Donna Hollis Hudson,
Mark Worrell		 sas batch, proc step source 6
33 An Economical Approach to Verifying the Accuracy of Keypunch Data Steven N. Blair,
Nancy N. Goodyear		 sample, interval variable source 6
34 An Interactive Interface to SAS/GRAPH for Constructing Charts Blair MacDougall With the increasing variety, features and low cost of graphics hardware. it is becoming critical that general user software be device independent. SAS/GRAPH provides this advantage. This paper describes a software frontend to SAS/GRAPH which allows users with access to an interactive graphics terminal with 1 ight pen, joystick or crosshairs to construct nonstandard SAS plots incorporating text, lines and shapes. This software generates a SAS/GRAPH procedure which constructs the plot on the desired device. Both SAS and nonSAS users will find this software to be a productivity enhancement tool since it permits the rapid development of word slides, flowcharts and Gantt charts - to name a few proc gmap, proc step source 5
35 An Overview of PROC TABULATE Alan Eaton,
Ken Howell		 proc tabulate, proc step source
36 Analysis of Covariance Using SAS GLM Shayle R. Searle Analysis of covariance methodology is re viewed, both from the traditional vieW]oint of an extension of analysis of variance and as a special form of linear mouel. SAS GLM output is illUBtrated for the analysis of covariance of completely randomized designs, randomized com plete blocks and latin squares, including both single elope and intra-class slope models ancova, covariance, proc genmod, anova, regression analysis, z-test source 10
37 Analysis of Time Interval Problems in Multiple Period Behavioral Treatments with Multivariate Measurements Clifford L. Hirsch,
Mary S. Hirsch,
Terri J. Rau		 Physiologic data from human subjects was collected for six dependent variables. There were four recording periods: baseline, pre-feed back vOluntary control, biofeedback training, and post-feedback vOluntary control. Each experi mental period was five minutes in duration and the physiologic data was printed at fifteen second intervals. SAS statements were used to transform the raw data and. then, SAS multivari ate analysis of variance with repeated measures design was used to determine an optional data recording time interval to be used in future re search on migraine headache pain response variable, anova source 2
38 Analysis of Two-period Crossover Trials During SAS Roger D. Aitchison,
Saul L. Gitomer,
M. Lorie Skalland		 proc step, wilcoxon test source 3
39 Analysis of Variance and Subset Selection S.J. Bean,
Paul N. Somerville		 sample, sample size, degrees of freedom, anova source 3
40 Analyzing Analysis of Covariance Models in Repeated Measures Designs Using SAS James E. Laughlin,
James S. Roberts		 The analysis of covariance in a univariate repeated measures context can be handled easily in PRoe GLM. The analysis can be of two basic forms. In one case, the covariate is essentially a between-subject measure, while the other possible situation involves obtaining several measurements on the covariate along with one or more of the within-subject factors. This paper will present a method in which PROe GLM can be used in a multi-step fashion to partition the sums of squares and perform the appropriate tests for all effects of interests in both types of designs proc genmod, covariance, repeated measures, regression analysis, proc glm source 4
41 Analyzing Congressional Career Patterns Using SAS Robert G. Brookshire,
Claudia C. Putnam		 This paper demonstrates the usefulness of SAS in solving unusual data problems, computing new variables, creating new data sets, and performing statistical analyses. The Inter-university Consortium for Political and Social Research (ICPSR) has compiled a data set containing biographical information on every member of both houses of Congress since 1789. This data set has a variety of uses for studying ~hanges in co~gressional career patterns. but it presents the researcher with a formidable set of methodological problems: 1) The unit of analysis in the ICPSR data is the individual congressman, while that used in congressional career research is typically the Congress as a whole. 2) The common measures of legislative professionalism must be calculated for each Congress proc step, regression analysis, missing value, proc nlin, correlation source 7
42 Analyzing Latin Square Designs via SAS James E. Laughlin,
James S. Roberts		 The analysis of research data arising from Latin-square designs has posed a problem for marry experimenters who rely on statistical programs to analyze their data.. Tht:! major reason seems to be due to the confounded VAriance components in volved in certain Latin-squares and a lack of an algorithm for dealing with such designs in the commonly used statistical packages. This paper will look at many of the frequently cited Latin square designs and also present a very general proce~~ for analyzing such data in PROC GLM. This method allows the analyst to partition the variability of all the effects in the design in cluding the variation due to any partially reflected interactions that may exist. Between-subject Designs One of the most appealing aspects of a Latin-square design is the ability to reduce a large, factorial experiment into a smaller number of cells proc glm, proc genmod, main effect, proc step, repeated measures source 4
43 Application of the Pareto Curves in Logistics Modeling Ross Dennis auc, normal distribution source 8
44 Automatic SAS Generation of Documentation for a Large Public-Use SAS Data Base Judith L. Cohen Social & Scientific Systems, Inc. has been, for 3 years, providing data processing support to the National Center for Health Services Research (NCHSR) of the Public Health Service in the editing, analysis, and presentation of data collected in NCHSR's National Medical Care Expenditures Survey. As analysis files are developed from the large SAS database, some will ultimately be made available for public use. Since the files are numerous and may contain up to 800 variables each, it became desirable to automate the production of detailed, error-free, easily-readable codebook documentation to accompany the data files as they are released to the public domain. This paper presents a system which was deve loped to meet this documentat ion requirement proc freq, proc step, dataset, sample source 6
45 Automation of Estimation and Sampling Error Computations Using PROCs NASSTIM and NASSVAR Greg Binzer,
David Morganstein		 This paper describes an estimation and sam pling.errors package for use with survey data consisting of two SAS procedures (NASSTIM and NASSVAR) and a SAS preprocessor. Survey esti mates and their associated sampling error sta tistics are computed for user~specified charac teristics using balanced repeated half-sample replications (BRR) of the full sample. The SAS preprocessor prepares the input survey data file for use by PROCS NASSTIM and NASSVAR. The preprocessor defines the half sample replicates to which an observation be longs and optionally incorporates user-specified ratio adjustments into "weight" factors which are permanently affixed to each survey obser vation. The procedures NASSTIM and NASSVAR perform the actual estimation and sampling error compu tations upon the survey data file. Both PROCS allow the user to specify arithmetic expressions to create "computed estimates" for which the full complement o[ sampling error statistics are produced sample, proc calis, proc step, binary variable source 6
46 Building a Network of Users for Education and Technical Support David R. Dolkart source 5
47 CAL/S-SAS Course Authoring Language P. L. Olympia macro flow control, fullstimer source 7
48 CAMTAB: A System for Producing Camera-Ready Tables with SAS Michael D. Rhoads While SAS procedures such as FREQ and MEANS are useful for producing simple descriptive statistics, the appearance of their output is not suitable for most publication standards. The CAMTAB system, on the other hand, enables camera-ready tables to be produced directly within SAS. This system eliminates the need for data to be transcribed from printouts and then retyped. The main elements of the CAMTAB system include a cataloged JCL procedure, a SAS FROC (LMSTAB) that summarizes the data, and a generalized SAS DATA STEP (TABPRINT) that does the actual formatting and printing. Counts, vertical and horizontal percents, means and medians may be produced and mixed within a single table. Among the other features offered by CAMTAB are subtotaling, selective underlining of portions of titles and footnotes, and rounding and formatting cell values proc step, dataset source 6
49 Comparing SAS to Traditional Programming Languages Claude Jackson proc step, proc print source 4
50 Computer-Customized SAS Graphics Robert Bury,
Claire Fromme		 A graphics package designed for production of finished output from high level commands can't be expected to anticipate every feature visualized by a user for a particular graph. Neverthe less, small changes in the standard product would often make a hi~hly acceptable graph out of an unuseable one. This paper describes computer proc step, proc sgplot source 6
51 Conditional Logit Analysis for Case Control or Employment Discrimination Studies Timothy Wyant conditional logistic regression, regression analysis source 5
52 Construction and Management of a Medical Research Data Base Using RAQL and SAS Jacob V. Aranda,
Ted Van Rossum		 relational database, normalization, structured data source
53 Cooperative SAS Usage: University of Toronto and Toronto General Hospital E.D. Mitchell,
Cathy P. Tong,
R.D. Weisel		 Toronto General Hospital is one of several teaching hospitals affiliated with the Univer sity of Toronto. The Cardiovascular Surgical Division of Toronto General Hospital uses the SAS package installed at the University of Tor onto in many of their research projects. With technical support from the SAS instal lation group at University of Toronto Computing Services 'U.T.C.S.), and the versatility of the SAS package, the division is able to per£ram sophisticated statistical analysis on its Metro Toronto adult op.en heart data. Monthly and qucq' terly reports on morbidity and mortality of pat ients undergoing routine surgery have been com piled and generated. The ability of the package to extract and analyse from a large volume of patient information collected at various inter~ vals helps in isolating factors associated with post-op morbidity and mortality data science source 2
54 Custom Report Writing Using SAS and the IBM 6670 Laser Printer Craig N. Mills,
David L. Ransen		 Computer-generated reports need not be dull and difficult to read. The IBM 6670 laser printer has considerable flexibility, hut. i~ ~um manly llsed for text processing. With the use of some simple techniques SAS can be made to produce high-quality reports that appear to have been professionally typeset. For applications wherein SAS output would normally be retyped to produce final drafts of reports, it is usually possible to eliminate t.hi~ step. thereby saving much time and completely eliminating transcription errors. Sample applications are presented below to demon strate the use of the DATA STEP and the PUT statement to control the operation of the IBM 6670 proc print, proc step, macro definition source 3
55 Data Base for Reporting on Patient Flow in a Clinical Group Practice of Medicine Ronald Anderson,
Susanne L. Daood,
Philip Reilly,
Priscilla Van Grevenhof		 source 4
56 Data Reduction and Summarization Joshua Sharlin proc step, proc means, cpu time, proc univariate, proc freq, proc sort, by group processing, record count, bar chart, sample, do loop source 8
57 DBase: An Efficient Interface Between the User and Large SAS Data Sets Russell Gilbert,
William Ingram III,
James W. Parkes		 dataset, proc step source
58 Demographic and Economic Information Applied to Business Decision Making Garry S. Meyer Donnelley Marketing Information Services Stamford, Connecticut With the release of data from the 1980 Census the demand for timely application of this information for management deci sion systems has rapidly increased. The work that we report on here involves the integration of virtually millions of in dividual statistics from both the 1970 and 1980 Census into a system which allows retrieval of informa tion for user d~fined areas of interest. These areas are constructed through a user-friendly interface which allows either existing geo-political areas to be selected and combined or geometric shapes defined (for example, a circle describes all the area within a fixed distance from a specified point). There are a wide range of applications both public and private sector, for such information. This paper will focus, however, on the utilization of this system for multi site consumer outlets where the areas defined approximate market areas sample, regression analysis source 8
59 Deterministic Inventory Modeling Using SAS Kent W. Bauman A user oriented interactive computer system has been recently developed using SAS to deterministically model long term inventory proje~tions for a multistage inventory to inventory production process. Alternative sce narios faT commodity supply and demand can be selected by the user to model the resultant inventories. Supply scenarios are further classified into known or committed sources and those that may be considered desirable sources of supply. One of two different models may be selected either to simulate inventory levels based purely on input data (descriptive model) or to generate unprescribed additions to inven tory (and the resulting inventories) needed to maintain existing policy requirements on minimum inventory levels (normative model). Inventory statistics are generated, printed and stored by a SAS background program based on the user's responses to several on line program input requests cpu time, data cleaning source 5
60 Development of an In-House SAS Basics Course Julie Stremel data science, dataset source 5
61 Discrepancy Analysis: Using SAS to Monitor Program Participation Richard J. Williford proc step, proc freq source 5
62 Du Pont SAS Applications Debra Milstein product profitability, data step source
63 EASYTAST: A Complete Data Management System Using SAS and TSO/SPF for Target Animal Safety and Toxicity Studies Federick E. Livesay,
Robert L. Van Duyn		 dataset, libname statement, data wrangling, response variable, work library, proc step source 4
64 Education/Consulting Panel Earl Nall,
Peter L. Rikard		 proc step, sample, sas consultant, proc genmod source 12
65 Effect Size and Power in Two Factor Nonorthogonal ANOVA: A Monte Carlo Investigation Margaret J. Martin,
William A. Powers		 anova, effect size, power analysis, sample, sample size, null hypothesis, monte carlo, main effect, proc freq source
66 Efficient Techniques for Large Data Files Judith H. Mopsik cpu time, sample, array, data step, conditional do loop, set statement, lookup, put statement, if-then-else source
67 Evaluating the Performance of VM/SP Systems with SAS (Presented as a "Birds of a Feather "session) Peter L. Jobusch proc step, dataset source
68 Evaluation of Non-accelerated Stability Data Kathleen Karpenter Dietrich,
Daniel L. Weiner		 A package of SAS macros has been written which allows users to evaluate the data acquired from a stability program. The package first reads the data from an ASCII disk file, prints descriptive reports, and creates graphs using SAS/GRAPH. Next, the packag2 fits several linear models to the data and determines which model fits best. If the chosen model is a straight line and if the slope is significantly different from zero, a shelf life will be estimated. Finally, a report is written which summarizes ~he results corresponding to the appropriate model estimate, standard deviation source 6
69 Experience with a Large Class Grade Decision Support System Eric Poon,
Marvin D. Troutt		 data step, discriminant analysis source 3
70 Experiences of a SAS Graphics User John L. Hall Jr. proc step, proc gchart source 3
71 FAA Aviation Forecast Error Analysis System Sivaram Kalyandrug,
Akira Kondo		 The Federal Aviation Administration publishes a number of forecasts of aviation activity for its various facilities: airports with FAA control to wers, air route traffic control centers, and fli ght service stations. The evaluation of forecast errors as actual data become available has been a primary concern of the Office of Aviation Policy and Plans, which is charged with producing and publishing the official FAA forecasts. Consequen tly, APO has implemented an FAA Aviation Forecast Error Analysis System by taking advantage of the many versatile features in the SAS data management techniques. This system has the following parts: 1. Sobering Diagram. 2. Actual Values and Yearly Forecast Series for Endogenous Variables. 3. Rela tive Forecast Errors (%) and Summary Report of the Latest Forecasts. 4. Relative Yearly Forecast Va riation. 5. Turning Point Analysis. 6. Elementary Forecast Error Statistics. 7. Theil Inequality Co efficient and Related Statistics. 8 residual, mse, dataset source 6
72 From Basic to Unusual SAS/GRAPH Applications C. Royce Claytor,
Nat H. Wooding Jr.		 proc step, proc gmap, proc gplot, proc gchart source 9
73 Front-End Wylbur Macros with SAS: Experiences in Faculty Workshops Alissa S. Bernholc,
Ronald P. Cody		 A series of Superwylbur macros has been written which allows SAS users to submit jobs without JCL and to retrieve SAS output and automatically report SAS and JCL errors. We have taught SAS workshops to our -faculty both before and after these macros were implemented. Their effect on novice users has been positive and dramatic, far more than we originally anticipated sas output, macro definition source 3
74 Generalized Program for Stratification &Strata Deterioration Analysis-GPSSD.. Bonnie Brown Jacobson sample, sample size, standard deviation, confidence interval, proc calis, proc step source 6
75 Generating an Image of an IBM System/38 Data Base in SAS Ben Burge,
Maureen Halloran,
C. David Hardison,
Robert Hizer		 This paper describes a system developed at Lilly Research Laboratories which provides utilities for producing all image of an IBM Syst_emj3& (S/38) data base in SAS. Image is defined as the data and its underlying definition within the S(38 environment as well as references to associated dictionaries_ The system includes methods for reading 5/38 file descriptions and automatically generating SAS code La build SAS datasets and format libraries dataset, proc step, format statement source 3
76 Graphic Comparison of Insect Populations Lawrence W. Grimes,
Hoke S. Hill		 proc step, proc gmap source 2
77 Graphic Presentation of Fisheries Radio Telemetry Data John M. Rinehart proc gplot, class variable source
78 Hierarchical File Generator by Non programmers Jack Gochenouer Large multinational corporations have a great deal of difficulty in consolidating financial information. This problem is com pounded in organizations which change manage ment unit reporting frequently. Harris Corporation has developed a system used both for production reports and ad hoc inquiries which allows for frequent change in accounting structures and is driven by nonprogrammers on the accounting staff. The key is the use of PRoe FORMAT and a SAS code generator. In an interactive session. an accountant builds a PROC FORMAT containing a hierarchy of accounts or corporate entities. A file or series of hierarchical files may then be built merely by specifying the row, column, and file (page) variables. The system then generates the SAS code, using the previously built PROC FORMATS. which will roll up the hierarchy and produce the report files. The system has eliminated dozens of manual reports, improved the response to changes in reporting require ments proc format, proc step source 3
79 Implementing On-Line Relational Data Bases with SAS Richard B. Waina SAS provides some good capabilities for implementing relational data bases in a TSO (Time-Sharing Option) environment on IBM/Amdahl computers. SAS has proven to be a flexible, easy-to-use and low-cost implementation with extremely broad data manipulation and analysis capabil ity. The data bases support selective inquiry using standard- or specially- fonnatted reports. Any desired subset of information can be produced in any required order by making minor run-time modifications of existing programs. For example, given two files, one containing coded Attributes of fiel d fail ures (e.g., ·location, date. type~ cause) and the other containing narrative information associated with each individual failure, a report can be produced containing all records satisfying a given selection criterion and sorted in a specified order, with the narrative infonnation properly attached to each failure record. SAS also provides nice facilities for summarizing and charting information proc step, histogram source 4
80 Increasing Productivity: Batch and On-Line-Interfaces to SAS Using SPF Dialog Manager Cary N. Prague sas output, batch job, error message source 6
81 Info-Menu Russell G. Hackett Jr. proc print, proc step source 7
82 Intelligent Financial Planning Models Using PROC LP Marc-David Cohen,
Alan Eaton,
William R. Gjertsen		 proc step source 6
83 Interactive Control Charts using SAS82 David C. Schlotzhauer sample, quality control, sample size source
84 Interactive Least Squares Estimation of Missing Values in any GLM Paul D. Hamilton,
Patric Dale McCray,
Michael C. Palmer		 missing value, proc genmod, proc step, proc glm source 4
85 Interactive Routines for Handling Time Series Data on SAS CMS John D. Friedmann time series analysis, proc step, time series data, data step source 5
86 Introduction to Linear Programming with SAS Marc-David Cohen proc step, dataset, sensitivity analysis, phase i source 8
87 Large System Implementation in SAS: The Resource Management System Using SAS, SASIFSP, SAS/GRAPH, and the Merrill Data Base Dennis M. Case,
Scott L. McGregor		 proc step, dataset source
88 LDMS: A Set of SAS Procedures for Longitudinal Data Management Philip S. Dow proc step, data dictionary, dataset, format statement, character variable source 6
89 Living with the AHA: The Construction of a Multi-Year Hospital Information Data Base Using the AHA Annual Survey and SAS Mireille Leger,
Frederick E. Pratter		 sample, if-then-else, missing value, time series analysis, do over, data cleaning source
90 Management Information Systems: An Environmetrics Approach Karen L. Daniels,
Michael P. Farrell,
Jonathan C. Goyert,
Rodney H. Strand		 data governance, management reporting, sample source 4
91 Missing Events in Survival Analysis Peter A. Lachenbruch,
Robert F. Woolson		 regression analysis, likelihood function, estimate, proc phreg source 6
92 Modes of Interactive Processing Cynthia Dietz,
J.D. Young		 proc step, proc options source
93 MULPOOLS: Analysis of Personnel Practices in Employment Discrimination Lawsuits Jon Chalfant-Walker,
Timothy Wyant		 proc freq, sample, exact test source
94 Multi-Time-Period NLIN Regression William J. Harrison regression analysis, correlation, predictor variable source
95 Multivariate Data Analysis Library: "French Methods" SYSPAD Jean-Michel Gautier proc step, dendrogram, proc princomp, proc corresp, data science source 6
96 Obtaining Part and Bipartial Canonical Correlation Estimates from PROC CANCORR: A Biomedical Application Steven N. Blair,
David A. Ludwig		 proc cancorr, correlation, proc step, proc genmod, proc glm source 3
97 Offense, Defense, and Causality in Baseball Charles F. Hofacker While baseball fans are known for their penchant for statistics, the application of rational statistical methods to the results of baseball games is apparently a new phenomenon (James, 1982). The lack of sophisticated analyses is all the more suprising when it is considered that almost every aspect of the sport has been religiously tabulated for over 100 years. SAS was used to enter and analyze published data fo~ tile American and National baseball leagues. Various econometric, time series, and simple descriptive techniques were used to look at a number of questions of interest to baseball fans, and perhaps those interested in the developement and structure of competitive systems. The first topic considered is the relationship between offense and defense. Is it indeed true, as Casey Stengel has said, that "good pitching beats good hitting and vice versa?" Other topics looked at include hitting, fielding, speed, pitching, and the relationship between the two major leagues correlation, time series analysis, regression analysis, moving average, logistic distribution, proc step source 5
98 Office Productivity Evaluation with SAS Nachum Finger,
Nelson M. Fraiman,
Sheldon Kutnick,
Murray Mohl		 proc step, anova source 5
99 On Estimating Spline Regressions Lawrence C. Marsh regression analysis, interest rate, proc step, spline, binary variable, maximum likelihood estimation, likelihood function source
100 On the Viability of Some ANOVA-based Mixed Model Estimates Kenneth L. Koonce,
W.G. Warren		 anova, likelihood function, proc mixed source 8
101 Overview of Categorical Data Analysis Methods Ingrid A. Amara,
Susan Atkinson,
Gary G. Koch,
William M. Stanish		 sample, covariance, weighted least squares, proc calis, goodness of fit, sample size, likelihood function, proc freq, maximum likelihood estimation, treatment effect, proc step, randomization, numeric functions, logistic distribution, chi-square distribution, permutation test, logistic regression, regression analysis source 10
102 PASCAL Speaks to SAS: A Simple Interface David I. Hopp,
Robert W. Staley		 clinical trial, proc format source 6
103 Percentile Matching: An Alternative Method of Curve Fitting James H. Smith Conventional methods of curve fitting, particu larly least squares, often have the undesirable property of destroying the a priori density of the predicted variable. That is, given a set of ordere~ pairs {(x:,y)} and ... an estimated function of x, f. the density of f(x) will in many cases di ffer markedly from the density of y. The histogram, predictor variable, correlation source 5
104 Power Comparison of the Weighted Squares of Means Method (Type III) and the Method of Fitting Constants (Type 11) in Unbalanced ANOVA Ramon C. Littell,
Richard O. Lynch		 main effect, proc freq, anova, harmean, proc genmod source 8
105 Practical Methods for Multiple Comparisons in MANOVA Keith E. Muller Multiple comparisons in univariate ANOVA have received continuous research interest for many years. Many practical methods exist. Powerful software has led to the common and increasing use of multivariate ANOVA. Al though the issue of multiple comparisons in MANOVA has received a great deal of theore tical interest, few practical results have been provided. Methods which can be computed are reviewed and evaluated. Issues include 1) whether the response variables are in a common metric, 2) the design of the study, 3) where to set type I error rate, 4) power, and 5) the type of comparisons of interest. Useful type i error, significance level, type ii error, anova, response variable, main effect, repeated measures source 4
106 Preparing Presentation Graphics with SAS/GRAPH Thomas J. Blashke,
Rodger Madison		 macro variable, bar chart, proc step source
107 PROC AID: A SAS Procedure Searching for the Best Model byStatistical Inference Yutaka Kimura,
Saburo Saeda,
Takashi Shitomi		 In marketing research,AID(Automatic Inter active Detector) analysis developed by Survey Research Center of Michigan University is wide ly used to identify clusters of observations having similar attributes under interaction. We have developed an AID algorithm together with SAS interface (AID/SAS). Our AID/~ao ~tart~; PROC AID; MODEL y= X1 x2 X3; RUN; Y dependent variable X1,X2 : independent variables Written in FORTRAN IV with uptodate function, AID/SAS is compact and portabl~. Cucc~nlly it runs on an IBM compatible machine.Facom M160 in our institute dataset, proc step, response variable, predictor variable, sample source 5
108 PROC DISCRETE-A Procedure for Fitting Discrete Probability Distributions Charles E. Gates,
James P. Geaghan,
George D. Williams		 poisson distribution, binomial distribution, proc freq, proc step, likelihood function, maximum likelihood estimation, negative binomial distribution, output, class variable, estimate, sample source 4
109 PROC MATRIX as an Instructional Tool in a Multivariate Statistics Course for Social Scientists Robert S. Barcikowski,
Randall R. Robey		 proc step, degrees of freedom, response variable, main effect, null hypothesis, proc genmod, regression analysis, multivariate analysis source 6
110 PROC MATRIX Macros for the Generalized Incomplete Model (GIM) and for the Multiple Design Matrix Model (MDM) Diane Fairclough,
Ronald W. Helms		 covariance, response variable, estimate, proc step, correlation source 6
111 PROC MCSTRAT Susanne L. Daood,
James M. Naessens,
Kenneth P. Offord		 predictor variable, regression analysis, relative risk, likelihood function, proc step, confidence interval, logistic regression, proc calis, binary variable, logistic distribution, covariance source 8
112 PROC Paired Bradley J. Dain,
Jon L. Kosanke,
Kenneth P. Offord		 wilcoxon signed-rank, normality test, sample, rank test, binomial distribution source 2
113 PROC PAIRED: Parametric and Non parametric Tests of Paired Observations Joseph R. Fischer Jr.,
William G. Jackson Jr.		 nonparametric test, proc step source 3
114 PROC SURVFIT and PROC SURVTEST: An Update Thomas R. Fleming,
Kenneth P. Offord		 life table, proc step, calendar year source 9
115 Processing Variable Length Records Peter L. Rikard length statement, dataset, trailing @ source 3
116 Program Strucure Analysis Using SAS Robert D. Lewis The Timesharing Secvices Department at the Minneapolis Star and Tribune Newspapers uses FORTRAN as its primary programming language. We have developed a SAS application that analyzes FORTRAN programs in terms of overall program structure. The program produces three outputs: (1) A matrix of CALLING to CALLED routines; (2) a chart showing the program hierarchy; and (3) a list ofroutines external to the program. While the program we use is restricted to analyzing FORTRAN programs, it could easily be modified to analyze any hierarchy that is defined syntactically in terms of owner/subordinate pairs. Thus the same method would be useful for the study of dominance hier archies in behavioral studies, or for analysis of reporting relation· ships in a large company. In our environment, this program is use· ful in that it reduces the time needed to document new programs or analyze old ones, thereby increasing the productivity of a small stair dataset, comment statement source 3
117 Randomization Tests Robert M. Hamer permutation test, randomization, correlation, sample, null hypothesis, anova, exact test, normal distribution, data analyst, one-way anova, fisher exact test source 9
118 RAQL-An Evolution in SAS Data Management David Burrage,
Michael Gilman		 5 AS enjoys widespread acceptance as a data management system. Most applications, however, involve the writing of sometimes lengthy and complex 5 AS programs. These programs are usually spec i fie to the data management problem at hand and do not easily generalize to other data management situations. The writing of ad hoc SAS programs to salve changing, often unforeseen data management problems results in heavy demands on programming staff and computer resources. The relCltiunal approach to the management of data provides a comprehensive and integrated solution to many data management problems. Although 5 AS currently has elements of a relational system, it most importantly lacks a systematized way to manipulate SAS data sets in a true relational manner. R AQL is a new relational query language which brings a full relational processing capability to 5 AS. The higher level language of RAQL permits the relational processing capability to be expressed clearly and very concisely relational database, data governance, macro facility, dataset source 4
119 Reading and Interpreting GLM's Printout of Estimable Functions and Sums of Squares Ramon C. Littell proc genmod, proc glm source 1
120 Reducing Maintenance Costs for SAS-based Systems Wayne Beekman,
Cary Toor		 data governance, proc print source 8
121 Regression of Simultaneous Systems Subject to Highly Nonlinear Constraints William Alexander,
C. Richard Shumway,
Hovav Talpaz		 E.tiaation of economic relationships subject to the theory typically requires imposition of both linear and nonlinear restraints. For ez ample. the behavior of a competitive profit m8%i_izinl producer can be represented by a profit function which is linear homogeneous and convex in product and input prices. The first derivatives with respect to prioes are the pro duct supply and input demand equations which are homoleneous of delree zero. Appro:zimatins the profit function by a second-order Taylor's e:z pansion in price ratios" the parameters of the system of supply and demand equationl compriae the Hessian matrix of the normalized profit function. The parameters are symmetric by Youna's theorem, and the Helsian is positive definite. Homogeneity is maiutained by normal ization, symmetry can be impoled by a set of linear restrictions, but convexity requires nonlinear restrictions. To maintain convexity. the problem il converted to a penalty-aulmented unconstrained minimization problem estimate, mse source 6
122 Robust Discriminant Analysis Anthony A. Billings,
E. James Harner		 Multivariate outliers often have pronounced effects on the interpre tations and conclusions of principal components and discriminant analyses. Their influence. however. is rarely examined due to the difficulty in dis covering them 1n high dimensional indi vidual space. This paper reviews Huber's tech niques for determining robust M-esti mates of mUltivariate location and dis persion. This allows principal compo nents to be computed which are less sensitive to outliers than those ob tained from the ordinary covariance matrix estimate. A robust discriminant analysis then is based on combining the location estimates from g groups into a between-groups covarinace matrix and the dispersion estimates into a within groups covariance matrix. The resulting generalized eigenvalue problem is solved to obtain the discriminants. A SAS program has been written using FROG MATRIX which allows classical or robust discriminant (and/or principal components) analyses to be selected discriminant analysis, covariance, sample, outlier, mahalanobis distance, proc step, robust estimator source 6
123 Robust Regression and PROC JACKREGR Daniel M. Chilko,
Gerald R. Hobbs		 regression analysis, degrees of freedom, residual, proc reg, sample, multiple regression, predictor variable, outlier source
124 SAS and APL: The Simultaneous Use of Both in Data Analysis Richard W. La Valley,
Eugene Mertz		 batch job, data science source
125 SAS and the Micro-in the Sports, Energy, and Financial Arenas R. Bryan Erb,
Gary E. Graybeal,
Cecil R. Hallum		 source 7
126 SAS DLI, SAS/FSP-Their Use in Large Scale DB Operations Richard H. LaRue,
Clayton D. Stewart		 account, input statement, macro definition, proc step, libname statement, sample source 6
127 SAS for Beginners Eileen Driscoll dataset, data governance source 6
128 SAS Macros for F-test Power Computations in Balanced Experimental Designs Andrew J. L. Cary correlation, repeated measures, intraclass correlation, degrees of freedom, main effect, sample, null hypothesis, probability distribution, sample size, macro definition source
129 SAS Methods for Balanced Repeated Replications Edward L. Spitznagel Modern sampling techniqup.s frequently involve complex strategies that are far removed from simple random sampling. Recent ievelopments in variance estimation take account of these complexities, but there is little in the way of generally available statistical software implementing Lhese sample, proc calis source 4
130 SAS Methods for Standard Error of Estimate in Maximum Likelihood Factor Analysis .. Susanna F.M. Clarkson,
Douglas B. Clarkson		 factor analysis, proc calis, factor loading, sample, dataset, covariance, by group processing, proc factor, monte carlo, proc step, macro definition source 5
131 SAS on a 32-Bit Mini Richard A. Usanis proc step, attrib statement source 7
132 SAS Tutorial: PROC PRINT Reporting James M. Watts proc print, proc step source 3
133 SAS Tutorial: Report Writing with FILE and PUT James M. Watts put statement, dataset, sas programming source 4
134 SAS, A Truly International Software Package? Akira Kondo SAS has become one of the most widely accepted data management and statistical analysis packages in the United States. Furthermore, SAS appears to be in the process of achieving a global acceptance as one of the premier software packages in the world. A recent issue of SAS Communications ( Summer, 1982 has listed SAS distributors in many parts o~ the world. For example, SAS has now over 200 products installed in Japan, a very important market in the global competition. This paper examines how SAS is thriving in Japan, a vastly heterogenous culture where a software package written in English has been accepted as a major tool in an applications package arsenal. There are somc formidable obstacles to be overcome between the two cultures. For example, SAS manuals must be translated into Japanese versions to make SAS generally accessible to Japanese users data governance, data science source 4
135 SAS/ETS: What Isn't and What Could Be Robert P. Parks proc step, proc reg, proc model, regression analysis, sample source 6
136 SAS: A Data Management Tool Generator Barbara Belasco,
Suzanne Ives		 In the clinical research environ ment, large amounts of information with different format and schemas must be processed easily and efficiently. Cer tain tools, such as data entry instruc tions, editing criteria, and data set creation and maintenance programs are usually created by the user. We have used UNIX and SAS for the past two years to generate these tools as well as secu rity and backup procedures. SAS genera tion pro9rams are written once and do not requ1re new programming effort to handle data with new formats or schemas. The use of generation programs provides increased efficiency and data set stan dardization. A commercially available data dic tionary has been implemented that con tains information essential to the pro cessing of clinical data and is used as input to the program generators. Other software, e.ga operating system files, also could be used as input to the gen eratorS dataset, data dictionary, sample, input statement, clinical data, proc format, libname statement, proc step source 6
137 SAS: A Decision Support System for Management Raymond A. Gibley,
Heidi J. Schoen		 In the past few years, there has been a grow ing controversy in the literature over the util ity of many Operations Research pub lica tions. Some readers argue, and rightfully so, that most of the art icles are authored by re searchers, statisticians, and mathematicians who have little knowledge of or regard for the true fabric of management. As such, mathe matical rigor and abstract thought are often emphasized to the exclusion of real world practicality. Consequently, the gap continues to widen between the academic theoretician who structures the methodology and the working manager who has to 1 i ve with the daily prob lems. To help bridge the gap, this paper focuses on SAS as a practical decision support system. At the outset, the authors would 1 ike to make ;t clear that there will be no appeal to rig or~ elegance~ or mathemat-ical gymnastics proc step, probability distribution, monte carlo, time series analysis, confidence interval source 6
138 Saving the Best for Last? First? In the Middle? Alternative Approaches to Teaching SAS and Beginning Data Processing David H. Christiansen,
James D. Hosking		 dataset, sas programming, libname statement, raw data, proc step source 4
139 Scheduling and Management of Periodic Reports Ross Merlin proc step, if-then-else source
140 Software for Statistical Graphics-An Overview Patricia M. Caporal,
C.J. Hahn		 anova, data science, regression analysis, statistical software, time series analysis, proc sgplot, correlation source 10
141 Some Future Development Directions for SAS Systems: Extensible Editors, Dynamic Graphics, Operations Research, and Automated Documentation Generators Scott L. McGregor,
Mary Nelson		 SAS provides a powerful implementation language for systems which are designed to be extensible, that is, where the users may extend and customize the capa bilities of the systems by writing and then executing new functions. SAS is easy to master and can handle high-level constructs. This makes SAS ideal for writing and executing extensions in real time. SAS 82 will further enhance this capability. This paper will investigate some areas where the extensible features of SAS may be put to greatest benefit. Suggestions for additional enhancements to SAS to further facilitate these kinds of systems will be made. In particular, the paper will focus on the use of SAS in (l) an extensible edi tor, (2) a dynamic graphics program, (3) an operations research system, and (4) an automated documentation generator. This paper assumes the existence of SAS, SAS/GRAPH and SAS/PSP in the working environment proc step, data step source 3
142 Some Guidelines for Documenting SAS Source Code Scott L. McGregor,
Mary Nelson		 Progrwmmers use a variety or techniques to document their programs. The styles. idioms and formats are frequently so diversiried that it is often difficult to maintain programs developed by several different individuals. Additionally, good documentation practices can ensure good progrrumming practices. This paper presents some guidelines for documenting SAS programs for easy mainte nance and comprehension. Guidelines are based on principles from diversified fields from systems design to graphics design. The guidelines discussed have been used by the author and several oth ers at the Hewlett-Packard Corporate Data Center with great success. Guidelines for Writing Easy to Read SAS Code and Documentation This paper is designed to provide SAS programmers with a guide to writing easy-to-read SAS code. It covers how to format raw code. choose names, decribe fields and provide additional documenta tion text proc step, _eof_ source 4
143 SPOTPLUS-Coal Market Data Management Robert M. Neil linear programming, dataset, sas log, temporary dataset source 3
144 Stratified Randomization Using SAS Joanne DiPietro,
Sampson Tu		 proc step, proc rank, proc sort, sas output, by group processing, proc print, randomization source
145 Structured Methods for SAS Applications J. Harvey Trimble Jr. data dictionary, error message source 4
146 Successful Statistical Consulting Douglas C. Montgomery The effective use of statistics is an impor tant part of the problem-solving process in in dustrial and engineering research. This paper identifies those aspects of the training and education of statistical consultants, that con tribute to their success in the industrial envi ronment. A number of issues regarding the professional practice of statistics are also discussed regression analysis, time series analysis source 4
147 Superimposing Wildlife Home Range Plots on Area Maps Lawrence W. Grimes,
Hoke S. Hill		 Frequently when radio-telemetry studies of animal location and movement are made, the re searcher wishes to display animal sighting loca tions, home range, and habitat utilization, and relate these to maps of the study area. Since it is virtually impossible to scale PROC GPLOT out put to exactly match the scaling of available area maps, the researcher needs to be able to create a map with the home range and habitat ut ilization superimposed on the study area map. This can be accomplished by digitizing the key features of an available map of the study area, giving a map dataset which can be merged with the animal sighting location and home range boundary point dataset. Then PROC GHAP can be used to generate a comprehensive choropleth map of the study area showing habitat types with home range characteristics superimposed proc step, proc gmap source 2
148 Teaching SAS as a First Computer Language Jean Ussery This paper discusses Introduction to Data Pro cessing Using SAS, the initial course in SAS/ Curriculum offered by SAS Institute. It de scribes the basic philosophy of the course, its design, content, and flow of instruction. The paper concludes with recommendations for any one developing a similar course dataset, raw data source 4
149 Testing Fourier Terms in Periodic Regressions John Jay Hilfiger Perlottlc phemoueua "re studied toy scientists in naoy fields Analysis ot coaplez cycles often Involves a representation by Fourier series and tests ot slanlllcane o± e&ch tern by means ol analysis of variance. The analysis of variance table provides a useful summary of the regreaalon results and helps the data analyst to select an appropriate aodel by deternlninj how Many terms should be included and whether a given model is a. good fit to the data* The SPECTRA procedure computes Fourier coefficients but does not provide an ANOVA table. This paper describes a set of SAS macros for ANOVA on results from the SPECTRA procedure and denonstrates Its use anova, response variable, regression analysis, proc freq source 4
150 The Development and Implementation of a Complete Forecasting System Entirely within SAS D. Benoit Sorhaindo,
U.V. Subba-Rao,
R. Suhir		 regression analysis, correlation source 4
151 The Information Center-A Review of the Concept Robert H. Torgler The Information Center is an organizational ap proach that offers potential productivity gains by effectively supporting end user computing to satisfy demand processing requirements. The Cen ter provides guidance, training, and assistance to end users in the use of appropriate products to improve the effectiveness and productivity of managers and staff profes:sionals, data wrangling, problem solving source 8
152 The SAS Macro Facility Joy Reel %if-%then-%else, macro facility, macro loop, macro put, macro definition, macro variable, proc step, proc print, sas log, macro call, title statement source 10
153 The SAS Macros for Non parametric Analysis of Repeated Measures Designs Melvin Moeschberger,
P. K. Tandon		 covariance, repeated measures, main effect source
154 The SAS Microcomputer Connection P. L. Olympia sample, bar chart source 5
155 The SAS Supervisor Donald J. Henderson dataset, keep statement, execution phase, proc step, sample, if-then-else, automatic variable source 8
156 The Use of Rank Analysis of Covariance in Analyzing Censored Survival Data Saeid B. Amini,
Peter A. Lachenbruch,
Robert F. Woolson		 A common problem in the analysis of censored failure-time data is the comparison of several treatment groups. Typically, covariables which are presumably related to the failure-time response must be accounted for in the desired treatment conparison. For the proportional hazards regression model, tests may be constructed from an appropriately specified partial likelihood, and computations may be performed using routines such as PHGLM. In certain situations only the covariate ranks are known, or the ranks are more appropriate than the actual covariate values for analysis. In these cases a modified rank analysis of covariance may be used to test the desired hypothesis. The requisite computations are also performed relatively easily using SAS. This rank covariance procedUre and the SAS computations are described in this paper; additionally, the SAS program for a complete example is included as an appendix ancova, covariance, regression analysis, likelihood function, null hypothesis, maximum likelihood estimation, proc phreg, clinical trial source 7
157 The Use of SAS in an Introduction to Information Processing Course Robert S. Workman An introduction to SAS is taught as part of the sophomore level Information processing course at Southern Connecticut State College. The students study a text with sections on computer technology. programming, systems, and computers in society. Interactive BASIC pro gramming is taught the first half of the term and SAS the second half. SAS is presented as an example of a modern problem oriented soft ware system. Its ease of use, good documentation, standardization, modularity, and continual updating are used to demon8~rate good programming practice sample, anova source 2
158 The Use of SAS in MIS-DSS Development David W. Johnson,
John R. Schultz,
Anthony A. Sinkula		 The terms Management Information System (MIS) and Decision Support System (DSS) have been used for severa 1 years. Recent ly, however, they have been given new emphasis with the advent of user-oriented database management systems (DBMS), personal work stations, computer graphics, and information management. This paper describes the basic components of MI S-DSS and discusses thei r benefits to an organization. The importance of a planned data science, mis source 6
159 Time Series Analysis of Highway Accident Data Ronald J. Classen,
David W. Elizandro,
Scott O. Starks		 This paper presents results obtained from the ap plication of SAS time series analysis techniques to fatal highway accident data fOL- the time peri od of ,Tanuary 1975 through December 1980. Special attention is given to data for the state of Ar kansas, although results from selected neighboring states are presented. During thie time period, many states, including Arkansas adopted programs incorporating countermeasures for the purpose of reducing the occurrence of alcohol involved ac cidents. This pap~r presents time series models for the preliminary analysis of the effectiveness of such programs. The paper contains background material for the problem considered, as well as a description of the data base and methodology used in the study. Motivation of study Any comprehensive program designed to decrease the number of fatal highway accidents must ad dress the problem of drinking and driving time series analysis, partial autocorrelation source 3
160 TITAN: Task Item Tracking and Nagging System Paul C. Willman sas programming, proc step source
161 Transferring and Processing a "Very Large "Data Set from a PDP 11/34 to an IBM 360/370 Keith E. Muller,
Kathleen J. Seiple		 adam dataset, final analysis, proc step, sample, proc sort, data governance, raw data source 6
162 Treating Multicollinearity with SAS William J. Wilson proc reg, proc step, regression analysis, correlation, sample source 6
163 Understanding the Methodology of the Construction of the American Hospital Association (AHA) Annual Survey of Hospitals; a Response to 'Living with the AHA: The Construction of a Multi-year Hospital Information Data Base Using the AHA Annual Survey and SAS' Roy A. Clark source 1
164 Univariate and Multivariate Repeated Measures Analysis Through PROCs REG, GLM, and MATRIX Robert S. Barcikowski,
Randall R. Robey		 repeated measures, proc step, multivariate analysis, covariance, degrees of freedom, proc genmod, proc reg, residual, sample, group by, univariate analysis source 6
165 Urban Spatial Structure and Gasoline Consumption Robert M. Ray sample, regression analysis, predictor variable, proc gmap source 6
166 Use of Principles of Composition in Selecting Type Fonts and Graph Types in Business Graphics Applications Scott L. McGregor,
Mary Nelson		 Business graphics systems such as SAS/GRAPH have made high quality graphics available to business professionals of varying levels. Many of these profession a1s have never had any graphics training, and yet in using the easily accessable new software they are preparing important graphics presentations. The persuasive ness of these graphics may be enhanced by use of basic graphics design principles. Some of these principles have been described in SUGI proceedings before (notabl¥. the selection and use of color was discussed at length at SUGI '82); however, principles of composition are rarely taught to business professionals who prepare business graphics. This paper will describe three elements of composition in ter.ms of three para digms and will illustrate their use ,n examples culled from the art world and as applied to Computer graphics source 4
167 Use of SAS/GRAPH in a Data Rich Organization: A Learning Experience Misa Gratton data science, proc step source
168 Using and InterpretingARIMA, X11, FORECAST, andSTATESPACE John C. Brocklebank This tutorial will examine a subset of those time !el'ies related procedures currently available to the user or SAS JETS release 82. Speci6cally by example, a diversi&ed group of techaiques will be introduced ranging from inex pensive, automatic rorecasting pr"edures, to more com plex, powedul, model based approaches. The variables examined in this discussion are ex traded. from -the CIT [BASE T AI: Citibank Economic Data Bue U5in~ the SAS procedure PROC CITmASE. CITIBASETAljs the economic data base distributed on magnetic tape by: Citibase/Citibank EeoDomic Department Citibank N .A. 399 Park Avenue New York. NY 10043. -1- These SAS PROCs perrorm the rollawinS tasks: FORECAST Fits univariate time series aDd automati cally produces forecasts. XU Seasonally adjusts moathly or quarterly time series usin8 an ada.pta.tioa of the U.S. Cense Bureau X·ll Seasonal Adjustment Program. ARIMA Analyzes aDd rorecasts uaiva.riate time series data., transfer runction da.ta, or intervention data proc step, time series analysis, proc forecast, regression analysis, proc esm, confidence interval source 8
169 Using PROC REG to Analyze Mixture Experiments Mark E. Showers In recent years, the use of mixture experiments has become a popular approach for developing or improving product formulations. A mixt~re experiment is used in applications where the response is a function of the proportions of the q components and not the total of the mixture. Mathematically this proc reg, proc step, regression analysis, multicollinearity, proc glm, proc genmod source 4
170 Using SAS for Data Communications Management G. Will Cook A data communications systems analyst needs to provide accurate, timely information to those who operate the network. A valuable but often overlooked source of data to help accomplish this is the systems source code which defines the network. This paper describes a SAS job stream which transforms this data into useful information. A SAS program is used to decompose the heirarchical data structure present in VTAM and NCP macros. From this, SAS files are created and information is disseminated through summaries and listings. On-line inquiry is available through SAS/FSP under TSO, and SAS/GRAPH can provide visuals. The job stream also passes SAS data to the Network Communications Control Facility (NCCF) for access through that system sample, title statement, if-then-else, proc step, format statement, put statement, macro definition source 6
171 Using SAS for Management of Research Resources Rosilind L. Ibrahim data cleaning, data governance, management reporting source 6
172 Using SAS to Compute Standardized Rates by the Direct Method Mary Anne Armstrong Standardization of rates is a frequently used descriptive technique in epidemiologic studies when certain factors, such as age, race, or sex, vary within the population of interest. A SAS program was written to compute age adjusted rates, by race and sex, for any categorical variable. PROC FREQ and PROC MATRIX were used to compute the rates, which were adjusted using the direct method, followed by a DATA STEP to format the output. The results are in crosstabular form, one cell for each combination of values of a classification variable and the categorical variable under study, including totals. The output includes, in each cell. the frequency, the crude rate, the adjusted stan dard deviation, and a 95% confidence interval for the adju~t~d rate proc step, proc freq, confidence interval, sample source 6
173 Using SAS with a DBMS to Form a Versatile Information System Stephen M. Smith SAS and a database management system (Model 204) have been linked to form an integrated information system called a Data Analysis System (DAS). DAS's have many~the processing capabilities that characterize traditional MIS's and Decision Support Systems (DSS). Like traditional MIS's, they are capable of fulfilling requests for rapid access and display of individual case data. In DAS's, however, the emphasis is on capabilities to perform ad hoc analyses across selected case data. Like DSS's. DASts~apable of fulfilling requests for sophisticated analyses. However, in DAS's the types of analyses are less limited, and can be performed with actual internal transaction data and a variety of external source data. SAS extends the analytical and graphical capabilities of Model 204, the core database management system. The SAS-Model 204 interface, EXTRACT/M, links these two components raw data, data dictionary, dataset, proc step, data cleaning source 6
174 Using SAS with RAQL in Data Base Courses David Burrage,
Michael Gilman		 The relational approach to data management is fundamental to university database courses. Unfortunately, very few relational systems are available at a price most universities could justify solely for teaching purposes. R AQL is a relational query language which operates in a SAS environment providing a low cost relational teaching system. The language of RAQL is based on the syntax used in the popular textbook by C.J. Date _ Introduction to Database Systems, Edition Two, Chapter 6. RAQL offers all the facilities in Date's book, and much more, and is implemented so as to M..Il at reasonable expense on any teaching database. Although no knowledge of SAS is necessary to use R AQL, it was found that the data management features of SAS reinforced and complemented the teaching of the relational approach. SAS used with RAQL at McGill University has completely transformed the undergraduate database courses dataset, relational database source 2
175 Using SAS/GRAPH to Produce Scheduling Charts Mike Atkinson,
Ron Boehm,
Sandy Hemphill		 phase iv, phase iii, phase ii, phase i source 6
176 Variance Estimation from Multi-Stage Sample Survey Data: The Jackknife Repeated Replicate Approach William D. Kalsbeek,
James Rochon		 sample, proc step, correlation, proc calis, regression analysis source 6
177 Viewing SAS as an End user Language Steven R. Schacht The continual increase in computing efficiency has created a large demand for new application programs. The backlog of requests for programming is growing in most companies. The best way to solve this problem is for end users to write many of their own programs in an end-user language. This paper examines several properties of good end-user languages and evaluates SAS against these criteria. Several areas where SAS can be made more user friendly are noted and the way the Federal Reserve Board is addressing these difficulties is discussed. The Problem In many organizations, the demand for computer services is growing at an extremely rapid rate. As machine costs decline (the cost per byte is cut in half every 18 months on the average), companies buy larger and larger computers. This new capacity should permit many more applications to be automated each year proc step, data wrangling, sample source 5
178 Within-Drug Changes in a Multicenter Clinical Trial and Their Applications .. Harji I. Patel change from baseline, proc step, ancova, sample size, proc glm, sample, covariance, proc genmod source

Back to Conf go 1984 go 1982

This site serves as a convenient hosting location and does not work as a publisher. All copyrights belong to their respective owners.

The information on this site is provided for general reference purposes only. Whilst every effort is made to ensure that the site is up to date and accurate, no warrant, nor any responsibility or liability for, the accuracy or completeness of the content or for any loss which may arise from reliance on information contained in this site is provided. Links to other sites are provided for your convenience but no responsibility or liability for the content of those sites or of any external site which links to this site is accepted. References to any specific commercial product, process, service, or corporation are also for information only, and do not constitute endorsement, recommendation or favoring.

SAS is a registered trademark of SAS Institute Inc. SAS, Statistical Analysis System, and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration.