I have been trying for several years to write a second book. I have a publisher who is willing to publish the book, but I can’t seem to find the time to write it. Just recently, I negotiated a change in my work schedule that allows me a bit more time for my own work, including this book, and I have been drafting up material on my website and my blog.
Blogs and webpages are not the same as book chapters, and the actual book will look quite a bit different than the individual pieces. But blogs and webpages are a start and they outline the particular topics that I want to address in a particular chapter.
Right now my working title is “So you’re thinking about research. How to make progress in any phase of your research”. I’ve used and rejected several other titles over time, so this working title will probably not survive either.
Each chapter in the book will cover a topic in research design, data analysis, or publication. For example, “So you’re thinking about a pilot study” is a chapter in research design and “So you’re thinking about writing your methods section” is a chapter in publication. For some of the chapters in design and analysis, I’m going to spend the first part of the chapter asking “Are you sure you really want to do this?” Many studies that are called pilot studies, for example, would be better off being described as something else, and many data analysis that might have started out thinking that linear regression was the best choice might be better off with a different type of analysis. Then I will outline some of the things that can help your get started in a particular area. Even within a narrow area like systematic overviews/meta-analysis, every research study is different. What I have found, however, is that the steps to get started are often the same. I will also include at or near the end of the analysis a section called “the fly in the ointment” that outlines some complications that you need to worry about but which are beyond the scope of this book. An example of this is an internal pilot study, which is a nice way of insuring that you can re-use the pilot data in the later study that the pilot was helping you plan. It’s a complication that requires that you do some more research and I offer some pointers.
If no dates are listed, presume that the work is from July 2016 or earlier. As I update and add material, I will probably add a brief comment like “new in September 2016.”
Earlier webpages that talk about my book are
- P.Mean: Jump start statistics for beginning researchers (created 2009-11-07)
- P.Mean: Jumpstart Statistics, a proposal for my second book (created 2010-07-23)
- P.Mean: Good news about my second book proposal (created 2011-01-01)
- P.Mean: Thinking about the title for my second book (created 2011-04-11)
I might want to start with an introductory chapter, where do research ideas come from?
Section 1: Research design
1.1 So you’re thinking about your research hypothesis. This is not written yet, but it is easy. I want to get you thinking about the outcome measure, the patient population, and the control population. The first section will ask if you really want or need a research hypothesis.
1.2 So you’re thinking about a pilot study. The current version was written in 2016 and is fairly polished and complete. There are several earlier versions that I should review before I start writing the official chapter.
1.3 So you’re thinking about quality improvement. This is not written yet.
1.4 So you’re thinking about a clinical trial. This is not written yet.
1.5 So you’re thinking about a chart review. The current version was written in 2016 and is mostly brand new material, so it may need a lot more polishing than other webpages or blog entries. But the topics are fairly complete. I am wondering if I want to talk about the electronic health record here or in a separate chapter.
1.6 So you’re thinking about secondary data analysis. This is not written yet.
1.7 So you’re thinking about a questionnaire. This is a very important chapter but I don’t have anything official for it just yet.
1.8 So you’re thinking about a systematic overview. The current version was written in 2016 and needs some work. Also, I want to split off some of this and create a new chapter on meta-analysis. The difference here is that the systematic overview chapter will really talk about the planning process before you actually get your hands on your papers and the meta-analysis chapter will talk about the statistical tools used to combine the information from the papers once you have them.
1.9 So you’re thinking about a power calculation. This was the chapter that I originally submitted in 2010 as part of my pitch to Cambridge University Press and others. I also have several other versions about power (actually sample size justification) that cover pretty much the same material. The structure might need a bit of changing to keep up with my recent concept of the book. For example, I need a section near the beginning that says, “Are you sure you want a power calculation” because power is only one of several ways that you can justify your sample size.
1.10 So you’re thinking about ethical approval. Nothing yet. I want to talk about studies that are not human subjects research and studies that are exempt. If you’ve followed most of the previous material in this section, you already have much of what you need to obtain ethical approval. Three important remaining issues are establishing your credentials as a competent researcher, demonstrating your concern for privacy rights, and preparing a defensible strategy for your data analysis. I’m not sure, but I also might want to tackle the risk/benefit ratio, though that may fit better under the section on power/sample size.
Section 2. Data analysis
2.1 So you’re thinking about data entry. This is sort of an in between chapter. It’s not really data analysis, but it fits here better than under research design.
2.2 So you’re thinking about a descriptive data analysis. Nothing yet. This is a foundational chapter for the remaining material, as you can’t run a t-test, regression, etc., without first running a descriptive data analysis.
2.3 So you’re thinking about a t-test. I have written quite a bit about this, but it is not available online just yet. This is a surprisingly difficult chapter to write because I have to address issues of ordinal data, lack of normality, and heteroscedascity. It may end up being too long a chapter when I am done, but it is easier to shorten a long chapter than expand a short chapter.
2.4 So you’re thinking about ANOVA. I have not started writing this, but it was a special request from one of the people I work with. It will be a hard chapter to write because it is such a broad area. I think I will limit it to single factor ANOVA and a two factor ANOVA with an interaction. It’s not clear to me if I should do this before or after the linear regression chapter. Perhaps I introduce a simple one factor ANOVA model here and talk about interactions after regression.
2.5 So you’re thinking about linear regression model. I don’t have a good prototype chapter yet, but it will be based on something I wrote way back in 1999.
2.6 So you’re thinking about a logistic regression. I don’t have a good prototype yet, but there is a lot of good material scattered throughout my website.
2.7 So you’re thinking about a survival model. I don’t have a good prototype yet, but there is a lot of good material scattered throughout my website.
2.8 So you’re thinking about non-linear regression. I got some feedback that this is too advanced a topic for this book, compared to the others, but I want to see if I can de-mystify the process. It is not as complicated (in my humble opinion) as you might think. But if we don’t have a chapter on this, I’m fine.
Section 3. Publication (includes presentation as well)
3.1 So you’re thinking about your abstract. This is not written yet. The abstract appears first, but should probably be written last.
3.2 So you’re thinking about your research overview. This is not written yet.
3.4 So you’re thinking about your results section. This is not written yet.
3.5 So you’re thinking about your figures. This is not written yet. This seems like it should be covered in the results section chapter, but there are some important issues with the arrangement and display of tables that makes this worthy of its own chapter.
3.6 So you’re thinking about your tables. This is not written yet. Again, this could be part of the results section chapter. I’ll know more about the organization once I start writing the results section chapter.
3.7 So you’re thinking about your discussion section. The current version was written in 2012, but is in pretty good shape.