What the evidence shows
This section covers what has actually been measured about these drugs, as distinct from what is claimed about them.
Most content about wakefulness and stimulant medication falls into two categories. Clinical sources report whether a drug is approved and broadly effective for its indication. Popular sources report what people say it does. Neither reliably reports the size of the effect, the conditions under which it was measured, or how confident the field is.
That gap is what this section addresses. Where a number exists, it is stated. Where the evidence is thin, that is stated too.
direct head to head efficacy trials between them
Weakly supported → firmly established
Cognitive enhancement: what the evidence shows
The research on modafinil and cognition in healthy people, including effect sizes, the domains where benefit appears, the domains where it does not, and the methodological problems the field has acknowledged.
Is modafinil a smart drug?
The term, what it implies, and whether the evidence supports it.
Research versus anecdote: what modafinil feels like
Why reported experience and measured performance diverge, and what that divergence means for anyone reading first hand accounts.
Wakefulness agents compared: the efficacy evidence
What the trials show for modafinil, armodafinil, solriamfetol and pitolisant, and why comparing them is harder than it looks.
Two ideas recur throughout this section and are worth stating once.
| The idea | What it means in practice |
|---|---|
| Statistical significance is not effect size | A finding can be reliably real and also small. A significant improvement means the result is unlikely to be chance. It says nothing about how much performance improved. Popular coverage usually reports only the first number. |
| Conditions determine results | Effects measured in sleep deprived people do not transfer to rested people. Effects measured in patients with a diagnosed disorder do not transfer to healthy volunteers. A result stripped of its conditions is not a result. |