Only a tiny percentage of the views expressed by the thousands of disabled people who took part in the government’s national disability survey have actually been read by a minister or civil servant. The vast majority were read, coded and analysed by a so-called ‘topic modelling’ machine!

The new information which came as a result of a freedom of information request, has added fresh ammunition for disabled campaigners who believe that the National Disability Strategy – which was “informed” by the survey results and was published in July – has no legitimacy and should be withdrawn.
Most of the questions posed by the UK Disability Survey in January were restricted to multiple choice answers. But four of them allowed “free text” answers, and the Cabinet Office says it received more than 25,000 answers from disabled people to these four questions.
A freedom of information response from the Cabinet Office now says that about 95% of these answers were analysed through so-called “topic modelling”, which the government has described as:
“a method of machine-assisted reading of text data, used to identify topics from free text responses to open format questions”.
A report on the survey responses says that only 1,200 of the 25,000 (4.8%) responses from disabled people were analysed by human researchers through “manual coding”, and even then, only with the aim of producing themes and sub-themes for the machine.
The freedom of information response says there is no written evidence to show how many responses were read in full by a civil servant, a minister or a researcher, with the Cabinet Office telling Disability News Service: “No information is held on the number of responses read in full by a Civil Servant, Research [sic], or Minister.”
The much-criticised survey is already being challenged by four disabled people through a high court judicial review.
Doug Paulley, one of the four taking the legal action, said:
“I am unsurprised but dismayed that the government evidently doesn’t care enough about disabled people’s input into their strategy that they didn’t even bother to read most responses. The topics chosen were not directed by disabled people, the mechanism of survey was inaccessible to many disabled people and the restricted range of answers meant that the limited free text responses were for many the only way to put down what is really important to them. I spent time writing mine carefully; doubtless other disabled people did similarly – unless they were put off responding altogether due to the other issues.”
This underlines the claim by the disabled people taking the government to court that the survey was so deeply flawed that the strategy has no legitimacy and should be withdrawn and rewritten with disabled people.
The Disability Unit have declined to say how the government justified having the vast majority of the free text responses “coded” by a machine rather than read by a human being. Although, let’s face it, this government haven’t yet shown that it has anything resembling a ‘human’ approach when dealing with disabled people!
Description of cartoon for those using screen reading software
A large grey Dalek is firing its laser gun at a document being held up by a wheelchair user. The document reads ‘feedback from disabled people’ and it is bursting into flames. On the floor around them are several more of these documents with burned edges. The Dalek has a sign upon it that reads ‘disability unit’ and is screeching at the disabled person: “Exterminate, exterminate!”
Posted by A6er on 28/10/2021 at 16:55
Reblogged this on Tory Britain! .