As indicated by the authors, this book is not intended to be used as a standard textbook. It is designed as a supplementary source or as a source for the researchers in the discipline. The book consists of 13 chapters covering several interesting topics in depth. Each chapter ends with a concluding remarks, references, and additional readings. There is no end of chapter exercise.
Chapter 1 covers the domain of social sciences, issues in this area, the role statistics play, and ends with an overview of the topics appearing in subsequent chapters.
The next chapter begins with the randomized response techniques (RRT) for quantitative and qualitative data and randomized response methodology (RRM). Extension of RRMs for two or more independent sensitive qualitative features of a population is presented next. The authors also discuss the importance of protection of confidentiality in collecting samples, which might result in increase cooperation by the respondents.
Chapter 3 deals with content analysis (CA) where the issue is to distinguish between qualitative and quantitative research. Three basic principles of a scientific method (objective, systematic, and generalizable) are presented. Through the application of CA researchers can review large volumes of data and explain the focus of individuals, group, or social attention. Other topics covered here are steps in CA and reliability of coded data. The chapter ends with limitations of CA including: (1) does not tell about casual conversations between variables under study and (2) cannot explain why certain trends emerge.
Scaling techniques are presented in Chapter 4. Here, the authors discuss the assignment of scores to responses to opinions or skills or competences or other attributes. In surveys, the respondents often have to select one of several possibilities such as strongly disagree, disagree, undecided or indifferent, agree, and strongly agree. In this chapter, we read different scaling techniques and other topics including scaling of test scores, percentile z- or sigma, t-scaling, and scaling of categorical responses.
In the next chapter, we read about data integration techniques where the idea is to rank the data based the degree they appeal and select the best and the worst ones in a multi-attribute environment. The topic of elementary methods for data integration starts the chapter. The authors then move to Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), which is a multi-criteria decision analysis method and present this issue in depth. TOPSIS is based on the concept that the chosen alternative should have the shortest and the longest geometric distances from the positive and negative ideal solutions, respectively. Among the topics covered here are elementary methods for data integration TOPSIS method along with an example.
In Chapter 6, the authors discuss the statistical assessment of agreement where judging quantities agreements among different experts or judges are presented. In this chapter, we see topics such as Cohen’s kappa coefficient and its generalization and the assessment of agreement in case of quantities responses.
Meta-analysis is covered in Chapter 7 where the idea is to make the use of all the available evidence which may be in the form of several pieces of information (derived from some data) or from different sources, some of which may be in the form of expert opinions. This would enable one to attach a weight to each piece of evidence. Topics include estimation of Bernoulli and several normal populations, and meta-analysis in regression models.
In Chapter 8, we read topics under clustering and discriminate analysis where under the first topic we put homogenous items in clusters. Topics include hierarchical clustering techniques, linkage measures, optimum number of clusters, partitioning clustering-k-mean method, classification, and discrimination followed by a detailed example.
Chapter 9 covers the well-known topic of principal component analysis (PCA) where the idea is to reduce the dimensional space of the problem where some variables (possibly a large number) are correlated or by considering only those variables which are actually responsible for the overall variations. That is, reducing the number of correlated variables while minimizing the loss. Topics covered in this chapter include the correlation vector diagram (biplot) and properties of the PCA.
The authors then turn attention to a related topic covered in Chapter 9, that is, factor analysis (FA). FA is also referred to as another way of reducing the covariates without losing too much of information. Factor analysis is related to PCA, but the two are different. Among the topics covered here, one learns about a number of rotations including varimax, quartimax, and promax.
Multidimensional scaling (MDS) is covered in Chapter 11 where the topic of scaling covered in Chapter 4 is extended to include multidimensional scaling. Types of MDS including non-metric, replicated, and weighted are also covered here. Additional topics presented in this chapter include MDS and factor analysis, distance matrix, goodness of fit, and matrix CMDS.
Chapter 12 covers the topic of social and occupational mobility addressing the changes or transitions of individuals across social groups or occupations where Markov chains and similar tools play an important role. Topics include possible measures of career patterns, career patterns based on Mahalanobis distance or entropy, along with an example.
The last chapter in the book addresses the social relationships among communities, social network analysis (SNA) and its features, making inferences about the network and how these features can be measured. Topics include sampling and inference in a SN, data structure in a random sample of units including some data types.
This is a good source for researchers in social science and related disciplines. It could also be used as a supplement in a related course. An introductory background in statistics should be sufficient to comprehend the topics covered in the book.
Saginaw Valley State University (Retired)