Instructor:

Gunnar Almgren

Office 127 L SSW, Phone 206-685-4077

e-mail: mukboy@u.washington.edu

Teaching Assistant

Chan-woo Kim

e-mail: chanwoo@u.washington.edu

Lectures Times:

Tuesdays and Thursdays 9:00-10:20 AM

SWS 125

Social Welfare 587 is the first quarter of a two-quarter required sequence in statistics for social welfare research offered to first year Ph.D. students. The basic objective of Social Welfare 587 is to establish a foundation in descriptive statistics, probability theory, bivariate descriptive and inferential statistics, and an introduction to multivariate correlation and regression. In addition. students will gain a fundamental understanding of the place, potential, and limitations of statistical analysis in the fields of social work and social welfare research. The basic perspective from which this course is taught, is that the statistical analysis of social issues and problems involves the responsible application of a set of mathematical concepts and tools in pursuit of ways to improve the human condition. Content will include ethical issues concerning the appropriate application, interpretation, and use of social research, as well as the potential limitations and biases of applications that fail to adequately consider issues of population diversity.

The approach of this course is basically conceptual, with heavy emphasis on the understanding of the logic of measurement and statistical inference. While the content includes the use of computer applications for the management of data and calculation of specific statistics, understanding of many of the critical concepts is best served by doing some hand calculation that involves use of algebra. A foundation in basic algebra is sufficient but essential preparation for this particular course.

By the end of the first quarter, students will:

1. Understand the relationship between statistical analysis and research design as applied to social welfare research.

2. Understand and be able to describe the general applications of descriptive statistics, inferential statistics, parametric statistics and nonparametric statistics.

3. Understand and be able to apply basic descriptive statistics to data in a manner that is consistent with scholarly convention. This includes the capacity to identify, calculate, and interpret the appropriate statistic across a wide variety of examples and data situations.

4. Understand the logic of statistical hypothesis testing, the types of errors that can be made, and the risksand trade-offs that are implicit within the hypothesis testing process.

5. Understand the components, calculation, application, and interpretation of basic inferential statistics to the analysis of relationships between variables.

6. Be able to use the personal computer version of SPSS to manage data and compute basic descriptive and inferential statistics, both to describe the relevant statistical characteristics of sample populations and to test hypothesis about associations among variables. This will involve both the identification of appropriate procedures and interpretation of computer output.

7. Be able to critically evaluate the use of basic descriptive and inferential statistics in survey research in published social welfare research.

8. Understand some of the ways in which statistics can be misused in social welfare research in a manner that perpetuates oppression and social disadvantage.

9. Develop a philosophy of honest and ethical use of statistics and reporting of research results.

1. Lectures will occur two days per week and will follow the textbook fairly closely. Weekly textbook and other readings need to be accomplished before class to promote mastery of the material. The teaching assistant will conduct a weekly review and question and answer session at a time to be arranged.

2. Homework problems are assigned each week, to be completed and turned in to the teaching assistant by class time Tuesday of the week following the assignment. Completion of homework assignments in most cases is essential to the mastery of concepts and exam preparation.

3. Weekly computer lab sessions will be conducted by the teaching assistant in order to teach and enhance the ability to conduct and interpret computer applications of statistical analysis. We will use SPSS-PC for Windows, available on the School of Social Work LAN. All lab assignments are to be handed in by class time Tuesday of the week following the assignment.

The lab assignments and some of the examples used in class will employ data from a subset of the NLSY (National Longitudinal Survey of Youth) utilized as the basis of Hernstein and Murray's arguments in The Bell Curve: Intelligence and Class Structure in American Life (New York: Free Press 1994). The structure of the file allows a broad range of inquiry into the relationships between child characteristics, parent characteristics, social class, and standardized measures of child development outcomes.

Learning Goals for Lab:

Students are comfortable with using SPSS 10.0 for analysis.

Students able to use window commands to run statistical procedures learned in class.

Students gain a working familiarity with use of syntax in SPSS.

Students able to select appropriate SPSS commands for a given analysis.

Students able to interpret output from analysis and to explain results in writing.

4. There will be two midterms and a final exam. The exams will be cumulative because the concepts

taught in the course are progressive in nature. Preparation for a cumulative exam is more likely to

promote long term mastery of key concepts.

5. The final grade for the course will be based on homework (20%), lab assignments (10%), midterm

exam 1 (15%), midterm exam 2 (25%), and the final exam (30%).

Required Texts:

Jaccard and Becker. (1997) Statistics for the Behavioral Sciences (Third Edition). Pacific Grove, Calif: Brooks/Cole.

 Bohnstedt and Knoke. (1994). Statistics for Social Data Analysis (Third Edition). Itasca, IL Peacock.

Reserve File: A reserve file of other required readings is maintained at the SSW library listed under the course.

We very deliberately selected a text that does a good job of explaining the critical content of the course in a manner comprehensible to students with a limited foundation in advanced mathematics. In some areas, students will benefit from a more in-depth discussion of the mathematical foundation and logic of the key concepts as it is presented in other texts. Sometimes students are better able to grasp the same material from another text because the manner of presentation fits better with the student's style of learning. For these reasons we have held some other texts on course reserve. You will no doubt note that these texts are dated, both in terms of publication and in some areas language and examples. However, they all provide some unique contributions to the understanding of the concepts taught in the course and in some instances have remained standard references for social scientists.

Carver and Nash. Doing Data Analysis with SPSS 10.0. (2000) Brooks Cole.

This resource is assist students who have not used SPSS with the opportunity to independently study data analysis with SPSS.

Blalock, H. M. Jr. (1979). Social Statistics, 2nd Ed. New York: McGraw-Hill.

Useful for linking mathematical reasoning to key statistical concepts, as well as a more in-depth conceptual discussion of many of the statistical procedures presented in class.

Blalock, H.M. Jr. (1982) Conceptualization and Measurement in the Social Sciences. Beverly Hills: Sage.

Good for thinking about conceptualization and measurement issues, and the link between them.

Siegel, Sidney (1956) Nonparametric Tests for the Behavioral Sciences. New York: McGraw Hill.

Older, but excellent reference text for nonparametric statistics.

The weekly schedule that follows is somewhat tentative, depending on the level of preparation demonstrated by the class during the early weeks of the quarter. The first 5 weeks of class covers topics and concepts generally found in any good introductory course to statistics, while the latter 5 weeks of the course covers material more typical to an intermediate level course in statistics. To the extent possible given the average level of baseline preparation of the incoming doctoral class, earlier introduction of intermediate level statistics will be pursued in order to gain a broader foundation in general linear methods by the end of the Soc WL 587-588 sequence.

Readings: J&B Chapters 1, 2

Reserve Readings: Walker. Degrees of Freedom. Journal of Educational Psychology 31 1940: 253-69

Homework:

Chapter 1: Problems 6, 7, 20, 26

Chapter 2: Problems 8, 13,17, 27

Probability and Statistical Inference

Essential Rules of Probability

Distributions Employed in Statistical Inference

The Central Limit Theorem and Extensions of the CLT

Readings: J&B Chapters 6, 7

Homework:

Chapter 6: 11, 12, 18, 27

Chapter 7: 1, 4, 16, 23, 25

Conventional Approaches to Hypothesis Testing

Type I and Type II Errors

Directional and Non-Directional Tests

Power Tests

The One Sample t-Test and Confidence Intervals

Readings: J&B Chapter 8

Weinbach, R.W. (1989). When is a statistical test meaningful? A practical perspective. Journal of Sociology and Social Welfare, 16 (1): 31-37.

Pillemer, D.B. (1991). One versus two-tailed tests in contemporary educational research. Educational

Researcher, 20: 13-17.

Homework:

Chapter 8: Problems 1, 9, 16, 17, 20, 22, 25, 27, 28

Analysis of Bivariate Relationships: Nominal Level Data

Readings: J&B Chapters 15

Homework:

Chapter 15: Problems 3, 17, 18, 19, 25, 26.

Independent Samples t-Tests

Correlated Groups t-Tests

Measures of Association with Nominal Level Independent Variables

One-Way Anova: Between Measures

Readings: J&B Chapters 10, 11, 12

Homework:

Chapter 10: Problems 1, 17, 19, 20, 22, 30

Chapter 11: Problems 12, 13, 14, 19, 20, 21, 22

Repeated Measures ANOVA

Measures of Association Derived from ANOVA

Two-Factor ANOVA

Main and Interaction Effects

Readings: J&B Chapters 12, 13, 17

Homework:

Chapter 12: Problems 14, 15, 20, 21, 22

Chapter 13: Problems 2, 5, 10, 17, 18, 19

Chapter 17: Problems 1, 3, 12

Bivariate Descriptive Correlation and Regression

Bivariate Inferential Correlation and Regression

Readings: J&B Chapters 5, 14

Homework:

Chapter 5: Problems 1, 10, 12, 21, 22, 23

Chapter 14: Problems 9, 18, 23, 24, 43

Midterm 2 November 20

Correlation and Regression: Two Independent Variable Models

Part Correlation, Partial Correlation, and Variance Partitioning

Readings: Bohrnstedt and Knoke, Chapter 8

Homework: B&K Chapter 8: Problems 1, 2, 4, 6 , 7

Analysis of Covariance

Ethical Issues in Research

Readings:

Review Session

The final exam for this course, according to the university wide exam schedule, would normally be Tuesday, December 18 at 10:30 AM. If all agree that an earlier exam date would be preferred, I am open to any exam date/time beginning December 13^th. We need to settle this before December 1, so a room for the exam can be scheduled.