Anda di halaman 1dari 2

S O C I A L S C I E N C E , S TAT I S T I C S & M A P P I N G

Recommended Reading for


SPSS & SAS Programmers
By Frank LoPresti
frank.lopresti@nyu.edu

S AS and SPSS are, along with


Stata, the most popular sta-
tistical packages at NYU.
These programming tools allow
researchers to analyze and manage
Ronald P. Cody and Jeffrey K.
Smith, Prentice Hall (referred to
as Applied Stats and SAS).

Learning SAS or SPSS is like


familiarizing you with the “look” of
SPSS. In those pages you are also
introduced to data preparation
through discussions of entering
data, sub-setting datasets, elim-
data. Additionally, SAS is used by learning how to fly. When you inating duplicate cases, and
corporations and universities as an meet your instructor, you can’t missing data concerns. The book
all-purpose programming environ- say “Don’t show me too much. goes into enough depth to make
ment to manage their information. I just want to fly to New Jersey you aware of the programming
Quantitative researchers also use once.” To simply survive, you must issues that a quick study guide
Stata, MiniTab, and S-Plus, along learn navigation, how to take off wouldn’t address.
with more exotic special purpose and land, how to relate to other For example, as shown in the
packages such as Lisrel, HLM, and planes, and so on. To complete a figure on p. 24, when introducing
OxMetrics. But most researchers simple analysis in SAS or SPSS, recoding (a process which might
start with SAS and SPSS. you must have programming and be used to create a categorical
The four books I suggest for data management skills, and know variable with two values [e.g.,
SAS and SPSS programmers are some statistics. Otherwise you’re US Born vs. Foreign Born] from
probably going to crash and burn. a categorical variable with many
• SPSS 12.0 Statistical Procedures Along with the skills that make you values, e.g., Country of Birth),
Companion, Marija J. Norusis, comfortable managing a dataset, Norusis shows how to check the
SPSS Inc. (referred to in this you must also have a sense of reasonableness of the result of
article as SPSS Stat Procedures); the process and where it is going; the recode by running a cross-
only then will the analysis prog- tabulation of the old variable,
• SPSS Base 12.0 Syntax Reference ress along fruitful paths. The books Country of Birth, against the new
Guide, SPSS Inc. (referred to as described in this article will help variable, US_born.
SPSS Syntax); you get started. Chapter 7, “Testing Hypotheses,”
is outstanding and will give the
• The Little SAS Book, A Primer
Primer, SPSS BOOKS beginning student of research
Second Edition, Lora D. Norusis’ SPSS Stat Procedures would methodology a strong under-
Delwiche and Susan J. be good subway reading for most standing of the reasoning behind
Slaughter, SAS Institute researchers, especially for those the procedures described in later
(referred to as Little SAS Book); returning to analysis after a hiatus. chapters of the book. If you took
The first 100 pages or so take you your last statistics course a couple
• Applied Statistics and the SAS through the functionality of the of years ago, this book is exactly
Programming Language, 4th Ed., various windows used by SPSS, what the doctor ordered.

Fall/Winter 2004 • Connect: Information Technology at NYU 23


Notice that the Brazil column counts 255 Brazilians as US Born—a contradiction.
Norusis advises the reader to run a cross-tabulation to check the new variable, US_born.

As a companion book to programs, which is followed by a Stats and SAS have chapter after
Norusis, I suggest SPSS Syntax, chapter-long syntax guide. chapter of detailed discussion
published by SPSS, Inc. It is Finally, Delwiche and Slaughter’s about a range of statistical proce-
important for several reasons. Little SAS Book demystifies SAS dures, such as frequency, descrip-
The “Universals” chapter, along through an inundation of small tive statistics, regression, repeated
with other sections, teaches examples. This 160-page book got measures, and other analysis
programming concerns that me over the hump in my under- of variance models. Discussion
many researchers never learn. standing of SAS. Over the years, topics include when a model is
Programming languages share I was in the habit of going to the appropriate, a look at underlying
certain core constructs, such as wall of SAS manuals we maintain assumptions, and details of where
data types, “if” statements, “do” at the ITS Social Science, Statistics and what the output shows us.
loops, and so on. SPSS Syntax pro- & Mapping Group offices at the ITS Both books go into detail on the
vides the depth of explanation that Third North Lab for advanced SAS many analysis of variance models,
an intermediate SPSS programmer programming concerns. SAS, to its sorting out issues like repeated
needs. For example, logical credit, has more manuals than the measures, unbalanced and bal-
expressions such as “and” and Little SAS Book has pages, and by anced design, missing data, and
“or” statements and “truth tables” consulting the advanced proce- interpretation of interactions.
are covered in this book. Although dures in these books, I was always There are hundreds of SAS
a thorough understanding of these able to address my needs for any and SPSS books on the market,
expressions is required for inter- given SAS procedure. quite a few of which I have read.
mediate-level programming, they When I started to teach the Many offer important contribu-
are often overlooked or avoided by introductory SAS classes offered tions to the subject, as they deal
non-programming researchers. by ITS, however, I read and reread with specific statistical areas such
the Little SAS Book to get conver- as time series or analysis of vari-
SAS BOOKS sant with introductory issues. ance. For beginning and returning
Cody and Smith’s Applied Stats Through this book’s deluge of short analysts, however, I recommend
and SAS is chock full of examples example programs, I came to an that they start with the four books
worked through to the results that understanding of SAS I didn’t have described here. Please also feel
show the output. The book also from using advanced procedures free to contact me to discuss
includes many problems for the or consulting the manuals as the training opportunities at NYU
reader to work on and even has need arose. Inspired by Delwiche (frank.lopresti@nyu.edu) or visit
an answer section. It concludes and Slaughter’s book, I wrote an http://www.nyu.edu/its/classes/ for
with seven chapters that teach SAS article for Connect last spring, enti- a list of free ITS classes.
programming: how to input data, tled “SAS for SPSS Programmers”,
merge datasets, arrays, character, describing my SAS catharsis Frank LoPresti heads the ITS
and numeric functions. These pro- (see http://www.nyu.edu/its/pubs/ Academic Computing Services’
gramming constructs are then used connect/spring03/lopresti_sas.html). Social Sciences, Statistics &
in a valuable chapter of example Both Norusis’ book and Applied Mapping Group.

24 Connect: Information Technology at NYU • Fall/Winter 2004

Anda mungkin juga menyukai