Reaction-time research is a crowded field with several genuinely good tools in it. Here is where DirectRT fits, and where we would send you somewhere else.
Most of the decision turns on this, and on whether your participants are in the room.
DirectRT does one thing: present a stimulus and record how fast someone responded, on a dedicated Windows PC, without asking you to write code. Lexical decision, sequential priming, evaluative categorisation, and similar paradigms are what it was built for and what it has been used for since 1997.
Choose an online platform instead if you need hundreds of participants from outside your institution. Browser-based timing has improved enormously and is now adequate for many designs — and no amount of lab precision helps if you cannot recruit the sample.
Choose MediaLab instead — or alongside — if what you are collecting is answers rather than latencies.
Compared on how each tool is built and paid for, rather than on a feature checklist — those change every release, and this does not.
| DirectRT | MediaLab | E-Prime | Inquisit | PsychoPy | Gorilla | |
|---|---|---|---|---|---|---|
| Where it runs | Lab PC, offline | Lab PC, offline | Lab PC | Lab PC or browser | Lab PC or browser | Browser |
| Built for | Reaction time | Questionnaires plus media | Experiments, broadly | Cognitive tasks | Experiments, broadly | Online experiments |
| Coding needed | None | None | Optional, Visual Basic | Its own script language | Optional, Python | None |
| Where data lives | Your machine | Your machine | Your machine | Your machine or their cloud | Your machine or Pavlovia | Vendor cloud |
| Pricing model | Buy once, own it | Buy once, own it | Per license, add-ons extra | Annual subscription | Free, open source | Per participant |
| Hardware integration | Keyboard, mouse, voice key, TTL | Keyboard, mouse | Extensive, via add-ons | Good | Extensive | Limited by the browser |
| Platform | Windows | Windows | Windows | Windows, macOS, web | Windows, macOS, Linux | Any |
Every product here is actively developed and we have tried to describe each one as its own makers would recognise. Details change; if something below has gone out of date, tell us and we will correct it. Product names are the trademarks of their respective owners.
Including, in every case, the situation where we would point you elsewhere.
Our own questionnaire and media program — the comparison researchers ask us about most.
You are collecting answers: scales, open-ended responses, consent and debriefing, or media presented as part of a session with branching between measures.
You are collecting latencies: how quickly a participant classified, categorised or recognised something, where the difference you care about may be a few dozen milliseconds.
Most labs run both. MediaLab handles the session around the task — consent, measures, instructions — and hands off to DirectRT for the timed portion. That is the pairing they were designed for, and the bundle costs less than two separate licenses. See how MediaLab compares →
The best known commercial package in cognitive psychology and neuroscience.
You need tight integration with EEG, fMRI or eye-tracking hardware, you are joining a lab already standardised on it, or you are replicating a study whose materials were written for it. Its timing performance is excellent and independently documented, and its position in neuroscience is well earned.
Your paradigm is a straightforward latency task and you would rather describe it in a spreadsheet than build it in a designer or script it. DirectRT is a narrower tool, bought once rather than licensed annually, with hardware add-ons that are not priced separately.
Commercial package strong in implicit measures, with lab and web delivery.
You want the same task to run both in the lab and online from one script, or you want to draw on its substantial library of ready-made published tasks — the IAT in particular. Running one instrument through both channels without rebuilding it is a real advantage.
Your work is lab-based and you prefer not to maintain a subscription or learn a scripting language. If your task is defined by a stimulus list and some conditions, DirectRT gets you there with less between you and the data.
Free, open-source, and genuinely excellent. We will not pretend otherwise.
Budget is the binding constraint, or you want open file formats, a very large community, and Python underneath. Independent timing research places it among the best performers available, in the lab and online. If you or your students are comfortable with code, it is hard to argue against.
You want a shorter path from stimulus list to running study, and a support line staffed by the people who wrote the software. Free is not free if it costs a graduate student three weeks — but if that student enjoys Python, it genuinely is free.
Browser-based platforms for remote data collection.
You need a large, diverse or non-local sample, or you simply cannot run people in person. Timing in the browser has improved to the point where independent studies find it adequate for many reaction-time designs, and a well-powered online study beats an underpowered lab one.
Your effect is small enough that measurement noise matters, or your participants come to you anyway. A machine you control removes the browser, the operating system and the network as sources of variance you cannot observe or report — and there is no per-participant charge as your sample grows.
Long-standing commercial package from Cedrus, often paired with their response hardware.
You are already invested in Cedrus response pads and want the tightest integration with them, or you need to run the same experiment on both Windows and macOS.
You are Windows-based and want your design to live in a spreadsheet you can diff, share with a collaborator, and generate programmatically from your own scripts.
Since this is the number everyone compares on.
You will see millisecond claims from every vendor here, ours included, and they are difficult to compare because they measure different things on different hardware. The most useful independent work is the timing mega-study by Bridges and colleagues (2020), which put a range of packages on the same measurement rig. DirectRT was not among the packages tested, so we cannot point you to a third-party number for it, and we would rather say so than imply otherwise. What we can tell you is that DirectRT synchronises presentation to the display refresh and ships with a built-in timing test you can run on your own machine — which, whatever software you choose, is the measurement that actually applies to your lab.
DirectRT is free for 30 days, with no payment details required. Run the timing test on your own hardware and see the numbers for your lab.