Explaining the process

It is difficult to analyse all of the research on Pilates, so we have started using AI.

At this stage, AI is not perfect, but it gives a good start. The process we go through:

1. Have the AI identify research around a particular topic. The application we are using only currently reviews 10 research papers, but they are generally from peer-reviewed, reputable journals.

2. If there are sufficient papers, it will generate a consensus meter like the one below.

There is a note with each reference as to the AI assessment of the research. It will say "Assessment" and then either:

• Positive (green)
• Possibly beneficial (amber)
• Negative (red)
• Unknown

3. We review the information that has been generated and edit, delete and add research as needed. If we do not agree with the AI assessment, there will be a note. For example: Assessment: Positive (green) (Amber in AI). This means the AI rated it as amber but on reviewing the article, it is clearly positive and so it has been changed.

Please note:

• the consensus meter is based on the original AI assessment. At this stage, we have not modified it based on our reviews.
• ​the assessment makes no evaluation of the quality of the research. If it is positive and included 10 subjects, it willl have the same rating as a positive study with 10,000 subjects.

4. For each article, there is a note regarding how influential the Journal is where the article was published. This will be a rating from 1 star to 4 stars, with 4 stars being the highest ranking. The measure used is the Scimago Journal Rank.

If a journal has a four star rating, the research can be given more weight than research appearing in a one star journal.

5. Each reference includes a brief summary of that particular article.

​Future enhancements

The aim is to get as many conditions on to the website as quickly as possible and then work to enhance the information. We will be advising on social media when a new summary is added to the site.

We are also working on a scoring system for the research as a systematic review should carry more weight than a pilot study.

​Stay tuned!