We all have a small number of people who heavily influence us. One of the big influences on my statistical thinking was Ray Thomas. I met him when we volunteered on several RSS committees. Sadly he died earlier this year. Links to some obituaries are below but his PhD thesis from 1999 is worth a read. The language and terms might be dated but “Statistics as facts about society” deals with many of the issues that plague us today.
Ray’s thesis (search for “Statistics as facts about society” available free of charge via the British Library’s Ethos system) is a comprehensive reading list of articles and books – essential reading if you want to understand the tensions in the official statistics system. It explains in a series of chapters:
- how statistics should reduce the area of uncertainty in policy making by “unobtrusively narrowing the area of disagreement”.
- that statistics have a policy affirming not a policy developing influence
- the doubtful value of performance indicators
- how statistical systems are or are not independent of management
- how the categorisations used in economic statistics invalidate the data as a measure of welfare
- the misleading nature of many local area statistics
- the issues around the “integrity” of official statistics
- why statisticians should be public, not civil, servants
- how the social sciences need to get into statistics
Pulled together in one place, this amounted to radical thinking at the time, rarely articulated.
I’m pleased to say that two full obituaries for Ray are now available: from The Royal Statistical Society and the Regional Studies Association. Others, including The Town and Country Planning Association and OUlife have been published but are only available to subscribers or members. Many thanks go to Dr David Webster, Honorary Senior Research Fellow at the University of Glasgow, for leading on these.
I first met Ray before widespread email (and certainly pre-social media) when a letter addressed to me, in spidery handwriting, arrived at the bank I was working for. On opening it I found a couple of my newspaper articles, torn out of newspapers, with annotations on them. The comments weren’t critical or nasty but they did suggest that I might possibly not have thought through what I was saying.
They came from Ray, with an offer of a drink next time he was in London. Thank heavens I met him as that evening started a long term relationship that changed my life. He taught me to think and to question – and, once you scratch below the surface, it’s funny how many statistics are not showing what they purport to show and what users believe they show.
Realising that group think based on such statistics isn’t always giving the right steer and then saying as much isn’t usually a convenient thing. Generally our statistical systems, along with most of our public (and private) bodies, much prefer – and spend great energy defending – the status quo. There’s no reward for those in the system is disrupt or question it. Any outsiders who see problems and dare to question are usually ignored as maverick outliers. It’s no surprise, most of us have been there at some point: it’s easier to ignore the complaint, and undermine the messenger, than do “stuff” better.
Ray wasn’t a forceful, public, pushy man but he had huge plausibility, was nearly always right and was a very decent chap. While he got little praise for his efforts, it’s funny, discussing his passing with senior statisticians from diverse specialisms, how many have had a story to tell of how he had given them a little nugget of insight or a prod that had changed their thinking. I am really pleased that his influence has been so wide and I have no doubt that the sum of his numerous little nudges have been to society’s benefit.