Background on strategy selection
We could categorise using tabs as a strategy for using a browser.
There is a lot of research on the acquisition and use of strategies.
My thesis (see repository page with pdf link) has an extensive literature review (see page 21 onwards). To quote myself:
Another line of research has examined how people use computer soft-
ware. A common observation is that users often plateau at strategies
that are asymptotically suboptimal (for a summary of this literature,
see Bhavnani & John, 2000; J. M. Carroll & Rosson, 1987; Charman &
Howes, 2003). Classic studies in time-and-motion found that industrial
workers rarely discovered the most efficient strategies for performing
their tasks (Gilbreth, 1911). Charman and Howes (2003) found that
about half the participants in their study shifted to a more
sophisticated graphics drawing strategy ... Charman
and Howes (2003) suggested that strategy shift may be greater in the
lab than in the real-world be- cause participants are focused more on
low-level task goals whereas real world workers are more concerned
with high-level project goals. Furthermore, any discussion of
optimality raises the question of what is being optimised, and the
various costs and benefits of searching for optimal strategies.
Similarly, Yechiam et al. (2004) examined the use of a mouse versus a script-based strategy, along with a much slower keyboard-based strat- egy, on a spreadsheet manipulation task. Half of the participants were allowed to use the mouse strategy from the start, while the other half were required to use the script strategy initially. The script was difficult to learn, but ultimately quicker than the other strategies. They found that participants were generally unlikely to switch to the script strat- egy when it was introduced half-way through practice. Also, although Yechiam et al. (2004) did not report individual-level results, the pattern of results suggested that the shift to the using the script-based strategy tended to occur on the first or second trial of its availability or not at all. It would also seem that the shift to the script-based strategy was abrupt.
Relating this to using tabs
Presumably you use tabs because you find that they improve your browser experience. And I can see an argument for saying that once you know how to use them, they are objectively useful.
However, that does not mean that failure to use tabs is irrational.
- A user may not be aware of the existence of tabs or may not know how to efficiently use tabs. Presumably there are a near infinite number of things that a person is not aware of, so a user may need a general decision rule to guide them as to when to explore features of an interface. In particular, if a user is not experiencing perceived problems, they may not seek out strategies such as tabs.
- Automaticity is a powerful thing. If a user has grown comfortable using a browser in a certain way, they can focus on what they need to get done. Being concerned about low level workflow such as using tabs can be disruptive. Thus, they may experience a temporary drop in performance as they try to learn how to use the new tool.
- People differ in the importance of computers in their lives. The less a user uses a browser and a computer in general, the less benefit they are likely to derive from trying to optimise their interactions with the system. The complexity of representing such a system also uses cognitive resources.
Many similar arguments can be made about other computer strategies (e.g., using shortcut keys, programming versus manual approaches, using text editing keys rather than mouse based approaches, using Vim rather than notepad, etc.). None of this denies the value of tabs or other efficiency boosting strategies. I think most power users of computers develop a repertoire of skills for exploring an interface and developing an efficient workflow. However, that which is optimal or natural for a power user, may not be such for a more casual computer user.