|
|
Therapist Directory: Find a Psychologist, Find a Therapist, Find a Marriage Counselor
PSYCHOLOGY TOPICS
Selected topics in psychology
and mental health.
|
|
|
|
|
THE THERAPIST PSYCHOLOGIST BOOK STORE
 | |
 View Larger |
Statistics for Social Data Analysis By David KnokeGeorge W. BohrnstedtAlisa Potter Mee ( Wadsworth Publishing )
Release Date: 2002-01-01
Average Customer Rating:
List Price: $145.95
Price: $131.35
Eligible for FREE Super Saver Shipping on orders over $25.
Availability: Usually ships in 24 hours
|  Add to Cart |
|
|
Product Description
The fourth edition of STATISTICS FOR SOCIAL DATA ANALYSIS continues to show students how to apply statistical methods to answer research questions in various fields. Throughout the text, the authors underscore the importance of formulating substantive hypotheses before attempting to analyze quantitative data. An important aspect of this text is its realistic, hands-on approach. Actual datasets are used in most examples, helping students understand and appreciate what goes into the research process. The book focuses on the continuous-discrete distinction in considering the level at which a variable is measured. Rather than dwelling on the four conventional levels-of-measurement distinctions, the authors discuss statistics for analyzing continuous and discrete variables separately and in combination.
|
|
 Statistics for indoctrination, philosophy of science for mediocrities
For academic philosophers of science sociology is not a paradigm of successful science. Earlier Bohrnstedt had enforced his ersatz philosophy of social science as editor of the journal Sociological Methods and Research. Now in this book, Statistics for Social Data Analysis, Bohrnstedt, Knoke and Mee attempt to indoctrinate students in this same ersatz philosophy of science.
The authors advocate their version of Haavelmo's "structural-equation" agenda, allege a distinction between unobserved conceptual variables and observable "indicators", and pontificate criteria for identifying causality prior to statistical modeling and empirical testing.
Contrast their views with some basics of contemporary pragmatism, which prevails in professional academic philosophy taught in universities today:
1. Pragmatist definition of "theory": A theory is any universally quantified statement proposed for testing. It is never defined in terms of any particular ontology - such as subjective motivations. Thus there is no philosophical problem of relating sociological theory to empirical model, because the theory is the model and the model is the theory.
2. Pragmatist criterion for criticism: Only empirical criteria may operate in the criticism of theories. Ontological ideas including preconceived claims about causality are never valid criteria. Thus theories/models may not be rejected merely because their equation specifications do not describe motivations, i.e. do not have a mentalistic ontology.
3. Pragmatist thesis of ontological relativity: The empirically tested and currently nonfalsified theory decides ontology including any claims about causality. Thus one does not firstly know causes and then make theories, but rather the empirically tested and nonfalsified theories/models describe the ontologies of their domains including causality.
4. Pragmatist thesis of pluralism: There may be and often are multiple empirically acceptable - i.e. tested and currently nonfalsified - theories/models. Thus they all make acceptably competing or complementary causal claims, so long as they are found to be empirically acceptable - i.e. not falsified.
In her book, History of Econometric Ideas, Mary S. Morgan writes that there are two ways in which econometrics has been used: (1) discovery or theory development and (2) empirical testing. The contemporary pragmatist philosophy of science assigns statistical analysis a fundamental role in theory development as well as in theory testing. Pragmatism thus invites use of data mining and artificial-intelligence computer systems, which can create and test literally billions of hypotheses.
I believe that this book, Statistics for Social Data Analysis, leaves the reader/student ignorant of the true capability of new technologies such as mechanized statistical analysis of social data for discovery, and that its provincial philosophy of science invites a Luddite attitude toward twenty-first century social science research.
Sociologists who are unaware of contemporary academic philosophy of science will likely not find this review helpful. More importantly such sociologists will also therefore be unable to exploit to their - or their students' - advantage the enabling freedom and contributing opportunities offered by the pragmatist philosophy.
For more: Google my book, History of Twentieth-Century Philosophy of Science at my web site philsci for free downloads, and to view my other book reviews at this Amazon site.
Thomas J. Hickey, Econometrician
|
 For students of social sciences ( cura_lee )
This book is a statistics textbook for students of social sciences, not high-end users. I read earlier edition of this book in undergraduate statistics course. In that course, only basics of statistics were instructed. In social sciences, they don't need to know A to Z of statistics for all they have to know is what the function of SPSS or SAS means and what kind of data is needed and how the data would be analyzed in the statistics packages. There is no need to derive the functions in the textbook mathematically as they do in the courses of statistics department. We should understand what the function means, not how it is derived. This book is written in this regard. Unlike orthodox statistics textbook, this book tackles only the meaning of the statistical methods. In doing so, this book illustrates the methods with various field works and SPSS exercises. This is the stance most textbook written for social scientists takes. It seems that this book succeed in achieving the goal. Explanations are succinct and examples are apposite.
But this book is not that useful when you should do real research. Most social sciences articles use more advanced methods than what this book introduces. This book is good enough to beginners, but not so to who would be real researcher. At that point, you should have read more advanced ones already. If not, you couldn't read a piece of article in the common journals.
|
|
|
|
|
|
Addiction
