A study from 16 years ago is hardly relevant anymore. Back in 2003, people were still familiar with Office 2003's layout; most people have long since forgotten that layout or never learnt it in the first place.

The author doesn't discuss users' existing familiarity with Office 2003 and they only mention the word 'training' once, that "software design to interact with technology should require the least amount of training as possible" whilst never acknowledging that training in, and even qualifications in, the use of the Office suite was very much a thing in the 1990s and early 2000s.

Even then, the most problems were had in Excel. Advanced usage of Excel is done by technical people who would have had some training. Word and PowerPoint weren't shown to have significant difference in usability; arguably, Word is the program most people forced to use the Office suite spend their time in.

Never mind the ways by which the Ribbon and computers have changed since Office 2007. Options moved around, the Ribbon height reduced, screens having gotten wider to compress fewer options into submenus…

The author states at the end of their conclusion:

> In order to determine if the result of the study with respect to the Excel 2007 application persists and are not due to the learning curve the experiment can be repeated with users having at least three years using this version.

Do you know if the author or anybody else followed up?

It would be more interesting to see a comparison between Office 365 now that the interface has effectively become the de facto standard (same as Windows, macOS, mobile, tablet, and the web version) and Google Sheets (which retains the menus, toolbar, etc.).

I'm no lover of the Ribbon myself but I feel like there's better evidence for it not being the ideal interface than this which wouldn't have convinced me even at the time.

This isn't the proof that'll bring down the titan.

The question is, if you grant that a different design is better in a vacuum, how to weigh that against the benefits of existing familiarity.

Except for all the people who use Google Docs, I suppose.

I would not be surprised to learn that substantial parts of the core of Office were rewritten to make that possible. Unlike Collabora/LibreOffice, Microsoft is one of the most well-resourced organizations in the world and can afford to do that kind of colossally expensive project. Of course, they'd need an extremely compelling reason to do so, but Google Docs was an existential threat to their market share.

Also, other commenters report that the real-time collaborative editing experience in Office is more sluggish than in Google Docs, and this is consistent with my own admittedly very limited anecdotal experience, and if this has persisted for years it may well be for deep architectual reasons.

Office for web and desktop office were literally separate teams, in separate locations, when I worked there. Complete separation unified only by a document output.

That is not going to happen with LibreOffice.