Big Data for Epidemiology: Applied Data Analysis Using National Health Surveys

Reviewed by Nancyruth Leibold, Chair and Professor, Southwest Minnesota State University on 2/15/23

Comprehensiveness rating: 5 see less

The textbook is comprehensive within reason. No book can be one resource that includes all. The textbook states the purpose is to help Master of Public Health (MPH) students use basic applied data analysis from national health surveys. The book is comprehensive for the stated purpose. It is a solid starting point. It is not all-inclusive of everything to know about big data. It is excellent for the indicated purpose.

Content Accuracy rating: 5

The textbook explains the steps of using common statistical tests and secondary data in SAS 9.4. The book would be helpful for those learning how to do projects using SAS. The book is for MPH students on basic applied data analysis, using data from publicly national health surveys. The content on how to use SAS 9.4 is correct. The content is accurate and practical for MPH students and other public health professionals. Chapter 3 includes literature reviews on the application of national data and the Health Information National Trends (HINTS) Case Study.

Relevance/Longevity rating: 5

The book includes four sections to introduce the student to national health surveys, basic applied data analysis, and more about common national health surveys, dissemination, and conclusion. There are 12 chapters in the book. The book includes relevant links to common national health surveys. Each chapter contains tables, links, references, and some include images, boxes, flowcharts, and step-by-step instructions.

Clarity rating: 5

The flow charts, boxes, images, and organization promote clarity. The step-by-step instructions and images encourage clarity. Some chapters have tables, boxes, and flowcharts that also promote clarity. The chapters are well organized. Chapter 11, Dissemination: includes great examples of posters for students to view. The practical information about the dissemination of scholarship in Chapter 11 is excellent.

Consistency rating: 4

The chapters are consistent throughout the entire book. Meticulous organization of the chapters is present. A consistent format is throughout the book. One improvement suggestion is adding terms at the start of each chapter and defining the terms for the readers.

Modularity rating: 5

The textbook includes smaller sections of reading and learning within the chapters. There is a good use of white space.

Organization/Structure/Flow rating: 5

Each chapter uses numbering and headings to organize the topics. Links are present and in working order. The textbook is optimal for people using a screen reader.

Interface rating: 5

There are no concerns with interfacing. The book is easy to navigate. No distortion of images or flowcharts is present.

Grammatical Errors rating: 5

The language includes proper spelling and grammar.

Cultural Relevance rating: 5

The author uses respectful and inclusive language and includes research studies with underrepresented groups. The research, including underrepresented groups, is appropriate, especially for a public health textbook.

Comments

Additionally, the book is accessible and available in many formats. Thank you for your work on this open-access resource!

Reviewed by Evidence Matangi, Assistant Professor, Taylor University on 11/11/22

Comprehensiveness rating: 3 see less

This is a very good high-level textbook; I am not sure what knowledge of Statistics is a prerequisite for the target public health students. I feel that there is a need for detail on say the characterization of Big Data, and difference with and to small data. There is a need for some literature on survey designs, types of biases, the challenge of confounding, study designs and how to address the limitations or their consequences on actionable decisions, data ethics and ethical reasoning are pivotal for data quality, timeliness, and analysis. How to address the aspects of coverage and response rate, in the cases where they are reported low and the need for representative data.

Content Accuracy rating: 5

For a high-level Statistics student, this is ideal for applications to personal data sets. The acknowledgement of the non-representativeness of the datasets, may put a challenge on EDI and stereotype effects on learning. But if this can be shrugged, either by pointing to other smaller representative datasets for practical or homework.

Relevance/Longevity rating: 5

I cherish the use of SAS, and the prospects for STATA, it would also be ideal to use R and acknowledge the active community updating it, with links for updates for software and the datasets.

Clarity rating: 5

Great clarity, though there is a need to clarify on the objectives, aims for the different datasets. In section 7.5 there is a need to explain how the weights are being determined.

Consistency rating: 4

one can easily flow from chapter to chapter, and the framework is seamless. Terminology needs to be consistent, exemplified by urban/rural then metro/non-metro. Some terms were used without elaboration such as causality, and Taylor-linear methods.

Modularity rating: 5

I cherish the consistency in modularity from chapter to chapter, it builds an anticipation of the work to be done.

Organization/Structure/Flow rating: 5

Excellent organization.

Interface rating: 5

This is a good practical manual interfaced with literature, way to go in our pursuit for active learning with real and relevant data.

Grammatical Errors rating: 5

Excellent grammar used here.

Cultural Relevance rating: 5

Text accommodates for EDI, the acknowledgement of lack of representativeness of the datasets and the efforts to address this through highlighting author's studies with minorities is a perfect salvage.

Comments

Covid-19 happened, an acknowledgement of it as a contextual/situational factor and how it may influence future investigations on such datasets may be a good plus, as students have or are already invested in Covid-19 research in public health.