Bafflegab

 

Back in 1983, when I started teaching writing, Americans frequently wrote sentences that baffled readers: a style called bafflegab.  Too many of us still do.

For example, today I was enticed by the title of an article about how I could teach better online.  But the article was so hard to grasp that I gave up.  

Here is a typical sentence of bafflegab:

Initially, ongoing survey data from 248 pre- and 47 post-program student surveys and additional summaries of course evaluations given to students within the MVCR program were analyzed for student beliefs about necessary qualities for their instructors. 

Why did that writer fail to put actions into verbs, so I could understand as quickly as I could read?  The writer preferred bafflegab, probably because it sounded impressive.  But bafflegab limits his discoveries to the elite few who have time and skill to decode them.

Why?  Is the purpose of research to “get published” and elude criticism through obscurity?  Is it true that “bafflegab pays,” as J. Scott Armstrong showed back in 1980? 

In the same source, I found other types of bafflegab:

 

Finally, design is an ongoing process.  Technologies, incoming faculty, student needs, curriculum, etc. are always changing.  The document will therefore continually need revision.  Document revision can be on a structured or unstructured time scale.  It can also be performed by any number of committee formulations.  Then the process of acceptance of any new formulation of the document must reiterate itself.

(If a camel is a horse designed by committee, imagine a horse designed by a “committee formulation”!)

Decoded, the paragraph says, "Update documents." Scarcely a discovery.

To heal those rapid-fire sentences, writing teachers tell us to “vary sentence structures.”  But arbitrary variation solves little.  The sentences above need to be combined, not arbitrarily, but so that the each one has a word or two, early on, that repeats from the previous sentence.  And on that foundation, each sentence’s end should add its new information. 

If sentences linked ideas, and if actions appeared not in nouns but in active verbs, writers would clarify what they meant.  And the results might truly impress people. But that style requires writers who know what they mean, and writers who prefer communicating over impressing.

Most people really do want to communicate.  They will give up bafflegab if they know how.

That’s why I teach.

Honest Statistical Writing

We’ve all heard that people can lie with statistics.  But if you write statistics, you do not wish to lie.  And we all dislike being lied to.

For those who write or evaluate statistical evidence, there is a handbook. It is called How to Report Statistics in Medicine, second edition, 2006, by Thomas Lang and Michelle Secic.  (It is published by the American College of Physicians).  The book is simply and clearly written, with examples.  It advises us generally about types of measurement and then covers kinds of evidence, such as

·      Various measures of risk

·      Hypothesis tests with their p values

·      Association and correlation

·      Regression and ANOVA

·      Surveys, case studies, cross-sectional analyses

Readers who don’t know statistics very well (as I do not) can use this book to judge statistical presentations.  This book would be ideal if you review articles for a science journal.

I learned a few things:

On page 3 -  Do not give people more than two significant figures.  That’s all they can grasp.  Do not write that in April 2011, unemployment was 9.4481% of the available U.S. workforce.  It’s clearer to say that, in April 2011, unemployment was 9.4% of …  

On page 22 - 

When I inform people about the risk (of what’s bad) or the probability (of what’s good), I need to tell them at least four things:

1. The numerator and denominator with their units of measurement (for example, 25 out of every 100 …).
2. The time span.  Am I reporting the probability that X will occur over one year?  Over 10 years?  Over the life cycle of a fruit fly?
3. The unit of population.  Every 1,000 live births?  Every 1,000,000 European residents?  Every 100 fruit flies in my sample?
4. The confidence interval (CI or p value).

When I read an odds ratio (page 27), I should remember that “4.2 times as likely as” does not really mean “4.2 times more likely than.” 

This book by Lang and Secic is helping me read better and listen better.

As you know, statistics often appear in tables or graphs.  If you need to create them, here is a basic tutorial: Including Tables and Figures in Academic Writing .  Its examples are not necessarily all in APA 6th edition style. You will want to check APA6 pages 116 and following.  But the facts are good.

The tutorial comes from our British friends at the Subject Centre for Languages, Linguistics and Area Studies, Avenue Campus, Highfield, University of Southampton, Southampton, UK SO17 1BF