# Statistics Program at MSU: 1977-1990

### Summary

In 1977, Jimmy Carter (U.S. President 1977-1981) was inaugurated. Senator Mike
Mansfield from Montana, who had been the U.S. Senate majority leader since 1961, retired
in 1977. During Carter's term, the nation was burdened by an energy crisis and very
high interest rates. The U.S. Department of Energy was created to deal with the energy
crisis. Beginning in 1979, Iran held 52 Americans hostage for 444 days in its U.S.
Embassy. The 1980 U.S. Census reported a population of 226.5 million, including 787,000
in Montana. In May, 1980, the Mount St. Helens volcano erupted in Washington; it deposited
ash on Bozeman causing the local schools to close because of hazardous air quality.

Ronald Wilson Reagan (U.S. President 1981-89) defeated Carter in the 1980 election.
A large cut in federal taxes in the early 1980s was followed by the transfer of responsibility
to states for support of various social services. Congress passed the Bayh-Dole Act
allowing universities to patent their discoveries that were made possible with federal
research dollars. President Reagan was re-elected in 1984.

Stock market indices fell sharply in 1987, creating an economic downturn. The
Yellowstone fires of 1988 burned almost 800,000 acres over five months; a heavy smoke
haze covered Bozeman. President George Herbert Walker Bush (41^{st} U.S. President 1989-1993), defeated Democrat Governor Michael Dukakis in the 1988
election. The first Global Positioning System (GPS) satellite was placed into orbit
in 1989,

In the world of statistics during this era, there was dramatic growth in research
and application of Bayesian methods which became computationally practical because
of new numerical integration schemes such as the Markov Chain Monte Carlo (MCMC) technique.
Schwartz published his Bayesian Information Criterion (BIC) for model selection. In
geostatistics, the variogram and kriging interpolation methods became popular. That
motivated associated research on other interpolation techniques such as smoothing
splines, generalized least squares fitting of polynomials, and Bayes linear statistics.
General linear models, including logistic regression, became conventional tools. In
general, statisticians were relying on regression models adapted for application to
non-Gaussian responses. Scatterplot smoothing, such as locally weighted scatter-plot
smoothing (LOWESS) and non-parametric regression became popular tools for viewing
data. Other statistical research interests included shrinkage estimation, empirical
Bayes analysis, the jackknife and bootstrap techniques, the projection pursuit technique,
and various meta-analysis techniques for combining data from several studies.

For statistical computations, STATA was released in 1985 and S-PLUS in 1988.
Ethernet, the first network to support 10 MB/s speed became available and networked
personal computers on local area networks became increasingly important. MSU had access
to supercomputers via the NSFNET network in the late 1980s.

The visionary collector George Keremedjiev founded the American Computer & Robotics
Museum in Bozeman in May, 1990. The ACRM is the oldest continually operating museum
of its kind in the world and an enlightening local resource for MSU statisticians.

This MSU era corresponds to the term of MSU President William J. Tietz (1977-1991).
The Fall, 1980, MSU enrollment was 10,745. In 1978, Montanan’s passed the 6-mill levy
for support of higher education. The proportion of the state’s general fund appropriated
for higher education in 1978 was only 2/3 of the proportion allocated in 1968.

Although MSU became greatly underfunded during the 1970s, the 1981 state legislature
increased funding for higher education. Tietz and John Jutila, VP for Research, set
out to make MSU a comprehensive research institution; the legislature and BOR supported
that goal. The Montana economy turned sour in the mid-1980s because of a floundering
natural resource industry and the stock market crash. For the next few years, the
university budget decreased, salaries were frozen, and new hiring prohibited. The
1987 Montana legislative session was a near disaster for MSU although Montanans passed
the 6 mill levy for higher education in the 1988 vote. By 1990, MSU was no longer
a "state-funded" university, it was a "state-supported" university. To maintain itself,
MSU was forced to impose a 45% tuition increase (inflation-adjusted) from 1980 to
1990.

The Tietz/Jutila vision, and the early 1980s support for research activities
led to a surge of research successes at MSU. By 1990. MSU was a recognized, regional
research university with a reputation for academic excellence. The MSU Fall enrollment
was 10,392 in 1990. That year, President Tietz announced his retirement.

Statistician Dr. Kenneth Jerome Tiahrt became the Department Head in 1977, replacing Bob Engle. In 1979 the name of the Department of Mathematics was changed
to the Department of Mathematical Sciences. In 1980 both the "MS in Statistics" and
the "PhD in Statistics" degrees were authorized by the Montana BOR. During this era,
the Statistics Program conferred 31 MS degrees and 6 PhD degrees.

By 1990, the number of tenure-track statisticians increased from four to nine.
For the academic program, this increase was illusory because many on the statistics
faculty were serving in other capacities. For example, by 1991, Ken Tiahrt was serving
as Department Head, Dick Lund was the Agriculture Experiment Station Statistician,
Marty Hamilton was the resident biostatistician in the Center for Biofilm Engineering,
and other statisticians also held split appointments. New statistics faculty hires
during this era included: Dr. Evan P. Smouse (1979-1983), Dr. Michael D. Huffman (1980-85),
Dr. E. Henry Lee (1981-1986), Dr. Bruce J. Collings (1981-1988), Dr. Ronald R. Christensen (1982-88), Dr. Robert J. Boik (1985-2009),* Dr. William F. Quimby (1986-2006), Dr. Jeffrey D. Banfield (1987-2012), Dr. Patricia
L. Munholland (1989 - 2003), and Dr. Richard J. Rossi (1989 - 1994). However, within
a few years of their arrivals, Evan Smouse, Mike Huffman, Henry Lee, Bruce Collings,
Ron Christensen, and Rick Rossi all left MSU to take positions elsewhere.*Pocket calculators were used for elementary statistics homework calculations in 1977.
Dick Lund's MSUSTAT package and commercial packages such as BMDP, were used in upper
division and graduate courses. The first mass-consumer desktop computers appeared
in 1977. When personal computers became available at MSU, Lund adapted MSUSTAT for
use on them. For a few years in the early 1980s, Lund's user-friendly package put
the Statistics Program on the leading edge of PC technology for performing calculations
by standard statistical methods.

Dr. Frank McFeely retired in 1985 after nearly 30 years in the Department. He was the first PhD in Statistics to serve MSU until retirement. Back in the 1950s, when he moved into supervision of the statistics program following Dr. Ostle’s departure, the statisticians crafted a strong, modern curriculum. As the first director of computing, he shepherded MSU into the computer age. Thousands of students learned statistics through his instruction.

In 1978, the statisticians offered 15 one-credit modules (10 class meetings over 2.5 weeks) on statistical methods topics at the upper-division undergraduate level. Four years later, Elementary Statistics for Business and a senior-level Time Series course were added to the curriculum . In 1984, the fifteen 1-credit modules on statistical methods were replaced with six 2-credit modules including two new topics of Statistical Computation and Exploratory Data Analysis. The graduate course on Multivariate Analysis was reorganized to include weighted least squares techniques for smoothing and analysis of discrete multivariate data.

In 1986, an additional requirement of a written report requiring about 90 hours of effort was added to the requirements for a statistics MS under the non-thesis option. The goal was to prepare students for the type of reporting expected in their profession.

The 1988-1991 Catalog was the last to offer courses under the quarter system. The BOR mandated that the university system convert to the semester system for the 1991-93 catalog.

**Detailed Chronicle of the Statistics Program: 1977-1990**

**Kenneth Tiahrt, Dept. Head (1977-93) circa 1990 photo**

On July 1, * 1977*, statistics professor

*, replacing Bob Engle.*

**Kenneth Jerome Tiahrt became the Department Head***.*

**Ken would serve in that capacity for 16 years**(1977-1993) after which he returned to teaching* In 1979 the name of the Department of Mathematics was changed to the Department of
Mathematical Sciences. *This new name acknowledged the fact that the Department housed three distinct programs:
Mathematics, Mathematics Education, and Statistics.

**In 1980 both the "MS in Statistics" and the "PhD in Statistics" degrees were authorized
by the Montana BO****R. **A graduate degree for statistics students would no longer be called a degree in "Mathematics
(Statistics Option)."

* In 1977, I (REL) accepted a 20% joint appointment to the Montana Agricultural Experiment
Station (MAES)*. I had been supplying statistical assistance to agricultural staff and graduate students
on an ad hoc basis for several years and now this service was formally recognized.
Dr. Ervin Smith, the full-time Station Statistician, had the responsibility of data
analysis using the main-frame computer. He employed several students and other staff
to punch cards, list and check data entry, and conduct multiple batch submissions
to analyze a data set. Most data processing was by special purpose programs, often
written by students over past years. Sometimes SPSS (Statistical Programs for the
Social Sciences) was used for more elaborate analysis. The Station Statistician’s
data processing service generally was offered for free to MAES researchers, but was
limited by time and budget. An experienced statistician collaborator was needed. I
had suitable experience and knowledge, and was willing to take the split appointment.

Marty Hamilton retained his 10% appointment to the WAMI program until 1985.

**Observations on Faculty Retention**

The Statistics Faculty encountered the same challenges and rewards as the general faculty. In this era MSU’s faculty benefited from a couple years of optimism followed by a longer period of morale damaging events, including hiring freezes, salary freezes, operating budget cuts, and Montana's political ambivalence about higher education. At the same time, Tietz and Tiahrt were asking the faculty to be more research-oriented while maintaining excellence in service and instruction. The academic environment was excellent, but the practical aspects were unsettling. Young faculty experienced stress because research success had recently become a critical issue when seeking tenure and promotion. Some faculty members felt they were being required to do more, but with less resources. Many faculty chose to ameliorate the lack of state funding by finding support elsewhere; e.g., outside funding sources via grants and contracts or non-teaching joint appointments to other programs on campus. During this era, there were many personnel changes. The turmoil of the times frequently caused promising young statisticians to leave MSU for more rewarding or stable positions, sometimes in another university, more often in industry or government.

Nevertheless, the Statistics Program made excellent progress. By the end of the period, it had expanded and improved the curriculum, was teaching many more students, and had increased research productivity. The number of tenure track statisticians increased from 4 to 9. The 5 new professors brought modern training and research areas that kept the curriculum and research program aligned with the profession. The increase from 5 to 9 statisticians is somewhat illusory. Many among the statistics faculty had taken joint appointments to other departments and were not full-time in Mathematical Sciences. For example, by 1991, Dick Lund was the Agriculture Experiment Station Statistician and Marty Hamilton was the resident biostatistician in the Center for Biofilm Engineering.

**— Marty Hamilton**

**The Interdisciplinary Activities of Statisticians Resulted in Administrative Complications**

The interdisciplinary activity of statisticians, an important characteristic of their discipline, has been practiced at MSU since the 1950s. But the increased emphasis on research in the 1970s and the emerging growth of the statistics faculty had exposed the need for administrative flexibility concerning statisticians’ academic positions. Between research grants and joint appointments, often only a fraction of a statistician's tenure-track duties was dedicated to teaching the Statistics curriculum. By 1980, joint appointments complicated planning, especially when predicting staffing needs. More statisticians were needed in Mathematical Sciences to cover the curriculum than would be the case if the statisticians were not working on collaborative research appointments, co-teaching courses in other departments, or engaging in other interdisciplinary activities.

Departmental decisions on teaching loads, tenure, promotion, and salaries were made more complicated by a statistician's interdisciplinary work. The University promulgated guidelines that required consideration of all contributions, not just contributions to the home department. However, MSU’s strong administrative lines flowing through the college and “home” department hampered practical application of the decision process for faculty who engaged in considerable inter-departmental and inter-college activities. As a result, the process sometimes led to a unique negotiated agreement between the faculty member and the relevant university administrators.

**— Marty Hamilton**

* 1978-1980 MSU Catalog *Statistics – Department of Mathematics, Head of Department: Dr. K. J. Tiahrt

Professors: K.J. Tiahrt; Associate Professors: M.A. Hamilton, R.E. Lund, F.S. McFeely;

Assistant Professors: R.D. Abbott (one year visitor).

This catalog listed neither the new statistics professors nor the newly-approved graduate degrees in Statistics.

For AY 1978-79, I (Marty Hamilton, MAH) was away from Bozeman on assignment to conduct
collaborative research in the Biometry Branch of the National Institute of Environmental
Health Sciences, Research Triangle Park, NC, under the auspices of an IPA (Intergovernmental
Personnel Act) grant. My MSU appointment remained unchanged, the university received
complete salary & fringe reimbursement from the grant, and I was relieved of most
departmental responsibilities. The grant also provided travel/relocation support for
my family. This arrangement freed my salary to hire a replacement biostatistician
for that academic year. Fortunately, * Dr. Robert D. Abbott* was available to fill in. Although permanent hiring was not possible at the time,
he would have been an excellent addition to the faculty. During his career, Abbott
authored 100+ papers and was elected Fellow of the American College of Epidemiology.
As of 2018, he was Professor Emeritus, University of Virginia in Charlottesville,
and actively conducting research at the Center for Epidemiologic Research in Asia
at Shiga University of Medical Science in Japan.

Beginning in September, 1979, I held a National Institutes of Health Research Career Development Award that ran for five years. That award funded my research during the summer and allowed me to teach one less course each academic year.

**Two new statistics faculty members**** joined the Department, Dr. Evan Smouse and Dr. Mike Huffman.**

* Dr. Evan Paul Smouse (1979-1983),* PhD in Mathematical Statistics from Oregon State University in 1979. His dissertation
“Bayesian Estimation of a Finite Population Total using Auxiliary Information in the
Presence of Nonresponse” was supervised by Dr. H. Dan Brunk.

* Dr. Michael David Huffman (1980-85)*, PhD in Mathematics from the California Institute of Technology in 1980. His dissertation “Efficient
Approximate Solutions to the Kiefer-Weiss Problem” was supervised by Dr. Gary Allen
Lorden.

**Graduate Degrees Awarded: 1978-1980**

One MS degree and two PhD degrees in statistics were awarded during this period.

* The MS degree was awarded in 1979 to the statistics student Roslyn Stone*. Ms. Stone went on to earn a PhD in Biomathematics at the University of Washington.

**The PhD in Statistics degree was awarded in 1978 to Melvin Gail Linnell, and in 1979
to Susan Marie Hinkins. **

- Dr. Linnell’s dissertation “Efficiency Evaluation for the Nested Cube Response Surface
Design” was supervised by Dr. Richard E. Lund.
^{3}

- Dr. Hinkins dissertation “Using incomplete multivariate data to simultaneously estimate
the means” was supervised by Dr. Martin A. Hamilton.
^{6}

In this 1978-80 Catalog, the Stat 216 course description stated “Desk or pocket calculators
are used for problem solving,” indicating that ** pocket calculators were readily available**. Otherwise,

*the statistics courses are unchanged from the previous catalog, with an important exception – the Statistical Methods courses.*

*To provide maximum flexibility for students when choosing what statistical methods
to learn, the statisticians partitioned the former three-course sequence (324-325-326) into
15 short modules of 1 credit each.* Each course met 10 times over 2.5 weeks. Click this link to navigate to the background
article on modular instruction at MSU: Modular Courses: 1977-1984

*(Refer to the following descriptions for quarter offered, credits, and prerequisites). Current statistical methods used in research. Problem assignments utilize the electronic computer with prepared software (MSUSTAT). Individual one-credit segments provide topic alternatives. Course scheduling format will enable registration for several segments in one quarter.*

**Stat 330-344 Statistical Methods**<MODULES>

*A, Su. I cr. Lect. PREREQ: 8 cr. math/stat.*

**Stat 330 Statistical Methods—Sampling, Estimation & Testing***A, Su. 1 cr. Lect. PREREQ: Stat 330 or equiv.*

**Stat 331 Statistical Methods—Comparison of Two Samples***A. 1 Cr. Lect. PREREQUISITE: Stat 330 or equiv.*

**Stat 332 Statistical Methods—Probability Modeling***A, S, Su. 1 Cr. Lect. PREREQUISITE: Stat 330 or equiv.*

**Stat 333 Statistical Methods—Attribute Data I***S. 1 Cr. Lect. PREREQUISITE: Stat 333.*

**Stat 334 Statistical Methods—Attribute Data II***A, W, Su. 1 cr. Lect. PREREQUISITE: Stat 330 or equiv.*

**Slat 335 Statistical Methods—Analysis of Variance I***W, Su. 1 cr. Lect. PREREQUISITE: Stat 335.*

**Stat 336 Statistical Methods—Analysis of Variance II***W. 1 cr. Lect. PREREQUISITE: Stat 336.*

**Stat 337 Statistical Methods—Analysis of Variance III***A, W, Su. 1 cr. Lect. PREREQ: Stat 330 or equiv.*

**Slat 338 Statistical Methods—Simple Linear Regression***W. 1 cr. Lect. PREREQUISITE: Stat 330 or equiv.*

**Stat 339 Statistical Methods—Correlation***W, Su. 1 cr. Lect. PREREQ: Stat 338 or Stat 339.*

**Stat 340 Statistical Methods—Multiple Regression I***W. 1 cr. Lect. PREREQUISITE: Stat 340.*

**Stat 341 Statistical Methods—Multiple Regression II***A, S, Su. 1 cr. Lect. PREREQUISITE: Stat 330 or equiv.*

**Stat 342 Statistical Methods—Non-parametric I****S. 1 cr. Lect. PREREQUISITE: Stat 342.**

*Stat 343 Statistical Methods—Non-parametric II***W. 1 cr. Lect. PREREQUISITES: Stat 331 & Stat 340.**

*Stat 344 Statistical Methods—Multivariate***1980-1982 MSU Catalog**

Statistics – Department of Mathematics, Head of Department: Dr. K. J. Tiahrt

Professors: K.J. Tiahrt; Associate Professors: M.A. Hamilton, R.E. Lund, F.S. McFeely;

Assistant Professors: M. D. Huffman, E. P. Smouse.

By 1981, among the 25 tenure-track faculty members in Mathematical Sciences, six were statisticians. However, three of those statisticians had appointments that took them away from statistics instruction; Tiahrt was the Department Head, Lund was the MAES statistician, and Hamilton had both partial appointments to other departments and academic year salary support from a research award.

The course list and descriptions in this catalog was identical to the previous catalog.

**I*** n 1981, Dick Lund accepted an appointment as the Montana Agriculture Experiment Station
(MAES) Statistician.* The previous MAES Statistician Ervin Smith retired that year. Dick Lund served as
his replacement on a temporary basis and, after a national search, he was selected
for the permanent position. Although being the MAES Statistician was a full-time job,
Dick occasionally taught a one-credit statistical mehtods course. He arranged a time
trade in which one of the statisticians, often Mike Huffman, would use the time thus
released to provide statistical support to MAES. This plan provided a teaching/research
balance for the professors. Lund also hired statistics graduate students for MAES
projects, essentially providing a Research Assistantship in lieu of a Teaching Assistantship.

**From Main Frame Computers to Desktop Computers: 1977-1990**

One of my (Dick Lund – REL) first projects as MAES Statistician was upgrading the research capability of the seven MAES Research Centers. Microcomputers and supporting software appeared on the market about that time (1981). Although many folks involved in MAES research and research management were doubtful that microcomputers were all that useful, nine were purchased – one for each Center, one for the Statistician, and one for the Dean of the College of Agriculture.

Ken Tiahrt acquired similar equipment for the Department of Mathematical Sciences, but one computer at a time. In the Department, an office was set aside as a computer room, initially for terminals connected to the main frame computer located in the basement of the Renne Library, then for the personal computers. Even after personal computers were installed in each faculty office, the computer room was maintained. The Department space was used by faculty and graduate students who used a sign-up sheet to reserve time on the equipment. Later the university set up desktop computer rooms around campus for undergraduate students. For many years, access to computers was provided in a room on the ground floor of Wilson hall beneath the Department of Mathematical Sciences.

The typical academic desktop computational resources in the early 1980s were similar to:

- Intertec Superbrain microcomputer, costing about $3500 (in early 1980s dollars) equipped
with a Z80 8-bit processor, 64KB RAM, two 5.25 inch 360KB diskette drives ;
^{9}

- DECwriter model LA-120 line printer ($2000);
- 110 baud acoustic modem for communication over telephone lines with other campus computers; and
- supporting software such as WordStar (for word processing), SpellStar, DataStar, SuperSort
from MicroPro International, SuperCalc spreadsheet program from Sorcium, and Dbase
II data-base management from Ashton-Tate .
^{12}

In 1981, the MSU Computer Center replaced its Sigma 7 computer with a Honeywell CP-6 which supported interactive computer terminals and enabled operation of all previous software previously run on the Sigma 7, including MSUSTAT. A Digital Equipment Corporation DEC12 VAX model 11/780 was also purchased that year as part of an upgrade in computer service.

Statistical analysis and statistical research calculations were often conducted using special purpose programs, mostly written in Fortran. The most popular mainframe statistical software packages were SPSS (Statistical Programs for the Social Sciences), which was available only of the CP6, and SAS–1 (Statistical Analysis System, version 1), which became available in 1983 on the VAX. The packages were maintained by the Computer Center. SPSS and SAS manuals were kept in the Math Department’s computer room. The packages had become standard analysis tools for many disciplines across campus. SPSS was favored by social science and education researchers. SAS was favored by biology and biomedical researchers. MAES researchers used both SPSS and SAS, as well as several special purpose packages (see the Era 8 topic "Statistics at MAES: 1981-1995").

The capability of desktop personal computers increased rapidly. The Department and the MAES had to upgrade the computers every 3 years of so. As the popularity of personal computers increased, statistical software had to keep pace. By 1987, BMDP, SAS and SPSS were being run on personal computers at MSU. These comprehensive packages had a long learning curve and for many data analyses MSUSTAT was the package of choice. However, the statistics faculty felt compelled to acquaint the statistics majors and graduate students with the BMDP, SAS, and SPSS which were the packages that the students probably would be required to use in their future jobs.

I continuously updated the MSUSTAT package so it would run on the latest personal
computer operating system. Personal computers allowed MAES scientists to perform statistical
analysis on site and MSU students could do their homework calculations in one of the
campus computer labs. I demonstrated MSUSTAT at the 15th Interface of Computer Science
and Statistics (1983) in Houston, Texas and I published two articles about MSUSTAT
in *The American Statistician*.^{15} By the late 1980s, the MSU Research and Development Institute was marketing MSUSTAT
and sold site licenses ($196) to various universities and companies, some in foreign
counties. MSU was available for free on campus and the manual was sold in the SUB
Bookstore.

**— Dick Lund**

**Graduate Degrees Awarded: 1980-1982**

During this period MS in Statistics degrees were awarded to six students and PhD in Statistics degrees were awarded to two students.

**MS in Statistics degrees were awarded in 1980 to Elaine Arrington (Beauchemin), Donald
Daly, Jay Strever, & Lawrence Swisher and in 1981 to Nancy Campbell (Albrecht) and
Grier Jefferis.**

**PhD in Statistics degrees were awarded in 1981 to James Stanley Bergum and Milton
Wayne Loyer.**

- Dr. Bergum’s dissertation “Simultaneous estimation of risk in several 2⨯2 contingency
tables : an empirical Bayes approach”;
^{18}was supervised by Dr. Martin A. Hamilton;

- Dr. Loyer’s dissertation “Using a serial dilution experiment to estimate the density
of organisms” .
^{21}was supervised by Dr. Martin A. Hamilton.

**1982-1984 MSU Catalog**

Statistics – Department of Mathematical Sciences – Head of Department: Dr. K. J. Tiahrt

Professors: M. A. Hamilton, R. E. Lund, K. J. Tiahrt; Associate Professors: F. S.
McFeely; Assistant Professors: B. J. Collings, M. D. Huffman, E. H. Lee, E. P. Smouse.

The catalog listed 8 statistics faculty members because * two new Assistant Professors of Statistics* joined the Department, Dr. Henry Lee and Dr. Bruce Collings.

* Dr. Edward Henry Lee (MSU: 1981-1986)* received his Ph.D. from Iowa State University in 1981. Henry’s dissertation “Estimation
of Seasonal Autoregressive Time Series” was supervised by Dr. Wayne A. Fuller.

* Dr. Bruce Jay Collings (MSU: 1981-1988)* received his Ph.D. in Statistics from the University of North Carolina in 1981. His
dissertation "The negative binomial distribution: An alternative to the Poisson" was
supervised by Dr. Barry H. Margolin.

From 1983 to 1988, Marty Hamilton co-taught Community Health and Environmental Health courses under a 15% joint appointment to the Department of Microbiology.

The 1982-84 catalog showed two changes to the statistics course offerings, a * new introductory course Stat 230 *for business students

*specific topic*

**and a***for the applied statistics*

**Time Series***Stat 427:*

**course****Stat 230 Elementary Statistics for Business** A, W, S, Su. 4, 4. 4, 4 cr. Lect. PREREQUISITES: Math 117 and an introductory course
in computing.

Presentation of data, measures of location and variability, probability, interval
estimation, tests of hypotheses for normal and binomial populations. Desk or pocket
calculators are used for problem solving.**Stat 427 Time Series** A. 4 cr. Lect. 3; Lab. 1. PREREQUISITES: Stat 336 and 341.

Statistical methods of time series analysis.

**Graduate Degrees Awarded: 1982 - 1984 **

During this period three students received the MS in Statistics degree and one student received the PhD in Statistics degree.

**MS in Statistics degrees were awarded in 1982 to David Baldwin, Suzanne (Herel) Strasser,
and Georgia Ziemba.**

**The PhD in Statistics degree was awarded in 1982 to Robert D. Chew. **

- Dr. Chew's dissertation “Estimating toxicity curves by fitting a compartment-based
model to median survival times” was supervised by Dr. Martin Hamilton.
^{24}

**1984-1986 MSU Catalog**

Statistics – Department of Mathematical Sciences, Head of Department: Dr. K.J. Tiahrt.

Professors: M.A. Hamilton, D.C. Taylor, K.J. Tiahrt; Associate Professors: R.E. Lund,
F.S. McFeely;

Assistant Professors: B.J. Collings, R.R. Christensen, M.D. Huffman, E. H. Lee.

There were 9 in the statistics faculty, including a couple of * additions to the Statistics faculty* during these years.

*joined the Department as an Assistant Professor of Statistics shortly after receiving a PhD in Statistics from the University of Minnesota-Minneapolis in 1983. His dissertation “Searching for the Lowest Price when the Unknown Distribution of Prices is Modeled with a Dirichlet Process” was supervised by Dr. Donald Arthur Berry.*

**Dr. Ronald Richard Christensen (1982-88)***, who specialized in functional analysis, decided to commit to statistics, teach undergraduate statistics courses, and become involved in statistics research. The field of robust statistical methods had shown the importance of statistical functionals and derivatives of functionals as measures of robustness. (Statistical functionals also were utilized by statisticians for the theoretical structure of bootstrap techniques.) Based on his commitment, Taylor was listed as a member of the statistics faculty.*

**Dr. Donald C. Taylor, Prof. of Mathematics*** Changes to the Stat course offerings* in the 1984-86 catalog included a new course Stat 280 – Special Topics, a

*(the module names and descriptions were revised), a*

**reorganization of the Statistical Methods modules in which the 15 one-credit modules were replaced with six 2-credit modules***, a*

**new one-credit module, Stat 345 – Statistical Computation***, the graduate sequence on Multivariate Analysis was reorganized, and*

**new two-credit module Stat 362 - Exploratory Data Analysis**

**Stat 630 – Recent Advances in Statistics was deleted.****Stat 280 Special Topics** (On demand). 1-5 Cr. Maximum 8 cr. Maximum of Stat 280 plus Stat 480, 16 cr.

Courses not required in any curriculum for which there is a particular need, or given
on a trial basis to determine demand. **Stat 345 Statistical Computation** W. 1 cr. Lect. PREREQUISITES: Stat 354, 356 arid 358.

Introduction to statistical computer software. **Stat 350 Statistical Methods: Sampling, Estimation & Testing** A, W, S. 2 cr. Lect.

Introduction to basic statistical methods for analyzing data which include summarization
of data, and statistical inference for one and two sample problems. **Stat 352 Statistical Methods: One-Way Analysis of Variance & Simple Regression** A, 5, Su. 2 cr. Lect. PREREQUISITE: Stat 350.

Simple linear regression model; least squares estimation; inference on slope, intercept
and expected response, analysis of variance table for regression; testing the model.
Calculations for a single factor analysis of variance model; orthogonal contrasts;
multiple comparisons; fixed vs. random effects; relationship with regression. **Stat 354 Statistical Methods: Analysis of Variance** W. 2 cr. Lect. PREREQUISITE: Stat 352.

Multifactor studies, analysis of 2-factor experiments, main effects, simple effects,
interaction; ANOVA table, F- tests, use of contrasts and multiple comparisons; completely
random designs, randomized complete block designs; problems with the model, testing
for block-treatment interaction, missing values; Latin square design, ANOVA; response
curves, orthogonal polynomials; unequal sample sizes in two factor experiments; random
effects models, mixed effects models; multi-factor studies. **Stat 356 Statistical Methods: Multiple Regression** W, Su. 2 cr. Lect. PREREQUISITE: Stat 352.

Regression, estimation, analysis of variance and inference using matrices; partial
F-tests, coefficients of determination, residual plots, data transformations, polynomial
regression, indicator variables and final model selection. **Stat 358 Statistical Methods: Attribute Data Analysis** A, S. 2 cr. Lect. PREREQUISITE: Stat 350.

Chi-squared tests for contingency tables; 2x2 tables, including Fisher’s exact test
and McNemar’s test; rxc tables, including residuals and partitioning; log linear models;
and higher order tables. **Stat 360 Statistical Methods: Nonparametric Statistics** S. 2 cr, Lect. PREREQUISITE: Stat 350.

Randomization tests; quantile test and confidence intervals for quantiles; one and
two sample Wilcoxon tests and associated confidence intervals; tolerance limits; other
sum-of-scores tests and confidence intervals; measures of correlation. **Stat 362 Statistical Methods: Exploratory Data Analysis (EDA)** W. 2 Cr. Lect. PREREQUISITE: Stat 352. Introduction to EDA, numeric summarization
of data, stem-and-leaf plots; other summary values (median. hinges), letter values,
re-expression to symmetry, validation of normality assumption; box-plots, outliers,
graphical display; unordered multiple batches of data, parallel graphic displays,
re-expression for stabilization of spread; ordered multiple batches (x,y) data, linear
fits and resistant lines, robustness; ladder of powers, re-expression to linearity;
time series data, trend line analysis; smoothing techniques. running medians, hanning.
search for patterns, residual analysis.**Stat 537-538-539 Multivariate Statistical Analysis** A, (Alternate years, will be offered 1985); W. (Alternate years, will be offered
1986); Su. 3, 3, 3 cr. Lect. PREREQUISITES: Stat 356, 409 for 537; Stat 537 for 538;
Stat 409, 538 for 539.

**Stat 537: Statistical methods for the analysis of continuous multivariate data**.**Stat 538: Clustering techniques, graphical and other exploratory methods for multivariate analysis.****Stat 539: Maximum likelihood and weighted least squares techniques for smoothing and analysis of discrete multivariate data.**

**Graduate Degrees Awarded: 1984-1986 **

During this period, six MS in Statistics degrees were awarded:

** to Lori Ellis, George Hembree, and Terri Welsh in 1984, and to Larry Blackwood, Christopher Edwards, and Anne Torbeyns in 1985.**

**1986-1988 MSU Catalog**

1984-86 Statistics – Department of Mathematical Sciences, Head of Department: Dr.
K.J. Tiahrt.

Professors: M.A. Hamilton, D.C. Taylor, K.J. Tiahrt, R.E. Lund; Assistant Professors:
R.J. Boik, B.J. Collings, R.R. Christensen

There were only 7 listed for the statistics faculty, because of some * important personnel changes*.

**Dr. Frank McFeely retired in 1985 after nearly 30 years in the Department. He is the
first PhD in Statistics to serve MSU until retirement.**** Back in the 1950s, he led the statistics program following Dr. Ostle’s departure.
He guided the program to a strong, modern curriculum, and as the first director of
computing, he shepherded MSU into the computer age. Thousands of students learned
statistics through his instruction.**

* Three statisticians joined the Department in 1985*, Dr. Robert Boik in 1985, Dr. Bill Quimby in 1986, and Dr. Jeff Banfield in 1987.

* Dr. Robert J. Boik (1985-2009)* received his PhD in Statistics from Temple University in 1984. Robert’s dissertation,
“Interaction in the Analysis of Variance: A Procedure for Interpretation and a Monte-Carlo
Comparison of Univariate and Multivariate Methods for Repeated Measures Designs” was
directed by Dr. Francis Chien-Yeh Hsuan. Dr. Boik arrived with significant knowledge
and experience in the behavioral sciences and multivariate analysis. His statistics
PhD was his second doctorate as he had previously received a PhD degree in Psychology
from Baylor University (1975).

* Dr. William F. Quimby (1986-2006)* received his PhD in Statistics from the University of Wyoming in 1986. Bill’s dissertation,
“Selected Topics in Spatial Statistical Analysis: Nonstationary Vector Kriging, Large
Scale Conditional Simulation of Three Dimensional Gaussian Random Fields, and Hypothesis
Testing in a Correlated Random Field” was directed by Dr. Leon Emry Borgman.

* Dr. Jeffrey D. Banfield (1987-2012)* received his PhD in Statistics from the University of Washington in 1987. Jeff’s
dissertation, “Constrained Cluster Analysis and Image Understanding” was directed
by Dr. Adrian E. Raftery.

Unfortunately for MSU, in * 1985-86, two statisticians left MSU for non-academic positions.* Both bright, personable colleagues had productive careers elsewhere,

**Dr. Mike Huffman (1980-85) at Boeing Corporation**

**and Dr. Henry Lee (1981-86) at the U.S. Environmental Protection Agency.****A written report requirement was added to MS program.** The Statistics Faculty began requiring non-thesis MS candidates to submit a written
report on a research topic or consultation. A faculty member would be chosen to mentor
each MS student writing project. This new requirement was designed to prepare students
for the type of reporting expected in their profession. The writing project was designed
to require effort equivalent to a graduate course of 2 credits (90 hours of work).
Students were encouraged to register for STAT 570 to earn such credit.

* Minor changes in the statistics course offerings* appeared in the 1986-88 catalog. Stat 408-09 were listed separately as Mathematical
Statistics I,II; Stat 505-06-07 were listed separately as Linear Statistical Models
I,II,III; Stat 537-38-39 were listed separately as Multivariate Analysis I,II, and
Discrete Multivariate Analysis; Stat 549-50-51 were listed separately as Statistical
Decision Theory, Statistical Hypothesis Testing, Statistical Estimation Theory.

**Stat 408 Mathematical Statistics I**W. 4 cr. Lect. PREREQUISITE: Stat 407.

Statistical estimation theory.

**Stat 409 Mathematical Statistics II**S. 4 cr. Lect. PREREQUISITE: Stat 408.

Statistical hypothesis testing.

**Stat 505 Linear Statistical Models I**A. 3 cr. Lect. PREREQUISITES: Math 333, Stat 356, 409, and 426. Projections, least squares analysis, distribution of quadratic forms, general linear models, estimation and testing for normal linear models.

**Stat 506 Linear Statistical Models II**W. 3 cr. Lect. PREREQUISITE: Stat 505.

Analysis of variance, regression analysis, analysis of covariance.

**Stat 507 Linear Statistical Models III**S. 3 cr. Lect. PREREQUISITE: Stat 506.

Residual analysis, model selection methods, colinearity, variance component estimation, mixed models.

**Stat 537 Multivariate Analysis I**A alternate years, will be offered A 1989. 3 cr Lect. PREREQUISITES: Math 333, Stat 356 and 409.

Multivariate normal and Wishart distribution, inference for mean vector (Hotelling’s T2), inference for dispersion matrices, multivariate graphical displays, robust estimation of means and dispersion matrices, cluster analysis.

**Stat 538 Multivariate Analysis II**W alternate years, will be offered W 1990. 3 cr. Lect. PREREQUISITE: Stat 537.

Principal component analysis, factor analysis multidimensional scaling, multivariate linear models, multivariate analysis of variance and covariance, discriminant analysis and classification, canonical analysis.

**Stat 539 Discrete Multivariate Analysis**S alternate years, will be offered S 1990. 3 cr. Lect. PREREQUISITE: Stat 506.

Maximum likelihood and weighted least squares estimation in models for discrete data, log linear models, linear logistic models.

**Stat 549 Statistical Decision Theory**A alternate years, will be offered A 1989. 3 cr. Lect. PREREQUISITE: Stat 409. Statistical decision theory, Bayesian analysis.

**Stat 550 Statistical Hypothesis Testing**W alternate years, will be offered W 1990. 3 cr. Lect. PREREQUISITE: Stat 549.

Statistical hypothesis testing.

**Stat 551 Statistical Estimation Theory**S alternate years, will be offered S 1990. 3 cr. Lect. PREREQUISITE: Stat 550.

Statistical estimation theory.

**Graduate Degrees Awarded: 1986-1988 **

During this period, five students were awarded the MS in Statistics degree.

*MS in Statistics degrees were awarded in 1986 to Cynthia Hertzler and Carole Jesse,
and in 1987 to Lisa Bogar, Robert Dracy, Jr., and Constance (Huckins) Welsh.*

**The 1988-1991 MSU Catalog**

Statistics – Department of Mathematical Sciences Head of Department: Dr. K.J. Tiahrt.

Professors: M.A. Hamilton, D.C. Taylor, K.J. Tiahrt.

Associate Professors: R.E. Lund, B.J. Collings,

Assistant Professors: J.D. Banfield, R.J. Boik, R.R. Christensen, W.E. Quimby.

There were 9 statistics faculty members during these years, however, there were personnel
changes in some positions. In **1988, Dr. Ron Christensen moved to the University of New Mexico****,**** Dr. Bruce Collings moved to Brigham Young University****,**** and Dr. Don Taylor returned to the mathematics group in the Department.**

* In 1989, two new assistant professors joined the statistics faculty, *Dr. Pat Munholland and Dr. Rick Rossi.

* Dr. Patricia L. Munholland (1989-2003)* received her PhD in Statistics from the University of Waterloo, Canada, in 1988.
Pat’s dissertation “Statistical Aspects of Field Studies on Insect Populations” was
supervised by Dr. John David Kalbfleisch.

* Dr. Richard J. Rossi (1989-1994)* received his PhD in Statistics from Oregon State University in 1988. Rick’s dissertation
“Nonparametric Density Estimation by Generalized Expansion Estimators -- A Cross-Validation
Approach” was supervised by Dr. Hugh Daniel Brunk.

**Statistics courses were identical to the previous catalog.**

### MSU converts to a semester system

This was the final catalog under the quarter system of instruction. * The next (1991-93) catalog lists the new semester-based statistics curriculum*.

As part of the conversion to semesters, the statistics faculty successfully petitioned
the Provost to require all research methods courses across campus that involve statistical
techniques have a prerequisite of a Stat-rubric course from the Department of Mathematical
Sciences. * The statisticians' goal was to eliminate existing duplication of introductory courses
and to produce a coherent path of courses across departments for students seeking
statistical training. *Thus the introductory statistics courses in Mathematical Sciences were restructured
to accommodate this expanded service load. The College of Engineering was the only
academic unit at MSU that refused to follow the Provost’s directive concerning statistics
instruction.

**Graduate Degrees Awarded: 1988 - 1991**

During this period, ten students received the MS In Statistics degree and one student received the PhD in Statistics degree.

*MS in Statistics degrees were awarded in 1988 to Joseph Cavanaugh, Mysore Gayathri,
and Penni Kolpin, and in 1989 to Jeannine Bennett, Pam (Johnson) Greteman, Charles
Todd, and Ann Walker, and in 1990 to John Steven Cherry, Ken Peterson, and Karen Summers.*

**The PhD in Statistics degree was awarded in 1988 to Stephen Daniel Kachman. **

- Dr. Kachman’s dissertation “Inference procedures for fixed effects in multivariate
mixed models” was supervised by Dr. RobertBoik.
^{27}

##### Stat Prog at MSU during Era 8

Next topic (Advances in Statistics) during Era 7

Table of Contents

Last revised: 2021-08-31