Multiple Regression with Discrete Dependent Variables by John G. Orme

By John G. Orme

So much social paintings researchers are conversant in linear regression concepts, that are rather user-friendly to behavior, interpret, and current. notwithstanding, linear regression isn't really applicable for discrete based variables, and social paintings examine usually employs those variables, concentrating on results resembling placement in foster care or no longer; point of severity of elder abuse or melancholy signs; or variety of reoffenses by means of juvenile delinquents within the yr following adjudication. This ebook offers distinctive discussions of regression versions which are acceptable for numerous discrete based variables. the most important demanding situations of such analyses lie within the non-linear relationships among self sufficient and established variables, and especially in analyzing and proposing findings. transparent language courses the reader in short via each one step of the research, utilizing SPSS and end result presentation to augment figuring out of the real hyperlink functionality. The booklet starts with a quick evaluation of linear regression; subsequent, the authors disguise easy binary logistic regression, which supplies a starting place for the opposite suggestions. specifically, comprehension of the hyperlink functionality is essential so one can later interpret those equipment' effects. notwithstanding the ebook assumes a uncomplicated realizing of linear regression, reports and definitions all through offer important reminders of vital phrases and their which means, and during the publication the authors offer unique examples in response to their very own facts, which readers may match via by means of having access to the information and output on significant other web site. Social paintings and different social sciences school, scholars, and researchers who have already got a easy figuring out of linear regression yet aren't as acquainted with the regression research of discrete established variables will locate this simple pocket consultant to be an awesome boon to their bookshelves. for extra assets, stopover at http://www.oup.com/us/pocketguides.

Show description

Read Online or Download Multiple Regression with Discrete Dependent Variables (Pocket Guides to Social Work Research Methods) PDF

Best social work books

Mentoring for Meaningful Results: Asset-Building Tips, Tools, and Activities for Youth and Adults

Mentoring will get a face carry during this guide for fostering a fit, profitable mentoring software. built with enter from enormous Brothers tremendous Sisters and MENTOR/The nationwide Mentoring Partnership, this advisor offers a accomplished procedure that elements within the wishes of the complete mentoring crew, together with application leaders, mentors, mentees, mom and dad, and caregivers.

Rape in America: A Reference Handbook (Contemporary World Issues)

This guide offers an outline of the ancient and social context of rape in the USA. • contains an in depth listing of firms that experience an important influence nationally or locally within the circulation to prevent and forestall sexual violence• presents targeted listings of print and nonprint assets

Therapeutic Communities: Past, Present and Future (Community, Culture and Change)

Analyzing the culture of healing groups, their ideas and their context, healing groups: earlier, current and destiny is a wide-ranging creation to either idea and perform. together with contributions through one of the most eminent figures operating in healing groups, the publication explores the inventive and unique techniques of such groups to psychiatric care.

Contemporary social work practice

An advent to the varied settings of social paintings, detailing the suitable coverage context, tools and techniques. summary: An advent to the various settings of social paintings, detailing the proper coverage context, equipment and methods

Extra info for Multiple Regression with Discrete Dependent Variables (Pocket Guides to Social Work Research Methods)

Example text

For example, information is often not provided in research reports as to how variables were chosen for the initial model, how variables were selected for the final model, and how effect modifiers and confounders were assessed for their roles in the final model. Without meaningful information about the modeling strategy used, it is difficult to assess the validity of the results. Introduction to Regression Modeling Moreover, you cannot divorce the issue of model specification from design issues. Issues related to model specification and interpretation are more difficult with non-experimental research, because we cannot be certain what other factors might be influencing the DV.

If we had omitted available time to foster from our regression model, we would have been left with a false impression about the effect of kinship care on the potential to foster challenging children. That is, the relationship between kinship care and potential to foster challenging children was confounded by available time to foster, and so the relationship was spurious. 06 when controlling for available time to foster. ). Linear regression assumes that the IVs and the error are independent in the population.

Statistics such as Cook’s D are examined to identify cases that exert undue influence on the estimation of the regression model. Comparable statistics are available for the regression models discussed in this book. We will discuss and illustrate these statistics and issues in subsequent chapters. Sample Size Sample size is an important determinant of the statistical power of tests of hypotheses and the precision with which parameters are estimated. ML estimation is known to work well with large samples.

Download PDF sample

Rated 4.68 of 5 – based on 49 votes