Saturday, December 13, 2008

Saturday Morning Musings - problems with text tables

From time to time something happens that make me realise (again) that I do not fit in with some features of the current world. The latest example is the constant use of text tables in official documents as a way of summarising and presenting information and desired outputs or results.

I, too, use text tables and they have their place. However, their use is now so wide-spread that it has become quite pernicious.

I have the good fortune to be a very fast reader. I also have a reasonably good memory for the things I read. Often in my work I have had to provide briefing or advice on documents - reports, statistical releases, plans, articles, books -  in very short time. My reading speed allows me to skim a document to get the guts very quickly, thus allowing me to move quickly to writing at least an initial brief.

You can actually see this in some of my blogging. I think of some pieces as briefing notes to my readers who (in this context) have taken the place of ministers or senior management.

The approach that I follow depends a little on my purpose and the nature of the document. However, in broad terms I start by skimming the table of contents. This tells me what's in the book or report.

I then look at the introduction or executive summary. This tells me what the writer thinks is important. Depending on purpose and the nature of the document, I may look at the index, using this to focus on specific areas.

With history books in my areas of interest, I actually use the index to determine the value of the book to me. If the index does not contain items that I consider to be important, or if on checking back to the relevant page I find the content limited or bad, I don't bother any more.

With official reports, I browse the report after looking at the introduction or executive summary. My purpose here is to flesh out but also check the material in the introduction or executive summary. I also look for particular points of interest, noting the page number in case I want to go back.

In most cases, I can now write my brief. In some cases, I may want to read the whole document in more detail.

I cannot follow this approach with text tables. Where the text table is important, I have to stop and work through it column by column, row by row. This is actually quite a slow process. However, problems do not end here.

By its nature, the English in the text table is shortened form, sometimes cryptic. This comes about because of the mechanical limitations associated with the use of tables. Only so many columns will fit on a page, while there are also practical limitations on the amount of text that can be included in a cell. All this makes comprehension harder.

Quite often, there is at least some disconnect between the text table and the document. This comes about because of the importance of text tables in setting out key recommendations and measurable outputs or outcomes.

Towards the end of the writing process, many official documents go through a committee process in which people focus on the table itself. Changes are made, but these are not always reflected in changes in text in the main document. Quite subtle shifts in wording mean that the main document and text table diverge.

Problems continue.

We live in a measurement age. When looking at the text table, people want to add measures, aka outcomes. The very structure of the table means that individual items are considered in isolation from the broader picture.

One of my colleagues described me as a framework person. I like to work out a general structure and then dig down.

From a policy perspective, the more policy objectives that you have, the greater the likelihood of conflict between objectives. Further, you are always dealing with a hierarchy of objectives, a cascade down. If you want to achieve a, then there are a whole lot of things that you need to do to achieve a. But these are all secondary.

In a measurement world, the manifold outcomes set out in text tables acquire a life of their own. They become the things that measure success independent of relevance.

My experience with successful policy development - and this applies in a business context as well - is that implementation is messy, requiring constant changes. Things are never what you expected.

The problem with the combination of measurement and text tables is that they lock policy into a rigid structure determined in advance of implementation. Once set, they continue independent of diverging reality.

Text tables are a useful tool - the matrix approach that Bob Quiggin and I talk about is an example. However, once they become the key way of presenting things, the end result is nearly always a mess.       

No comments: