Why this matters
If your typing test measures only pure words‑per‑minute, it misses where professionals lose the most time: the little trips between keys and cursor. Every switch from typing to pointing and back is “homing” — and it adds up. The idea behind integrated pointers (TrackPoint‑style sticks, roller‑bars, and the new center‑touchpad keyboards) is simple: reduce homing so you stay on home row and in flow.
KLM (the Keystroke‑Level Model) assigns a standard 0.40 s cost to each homing action H — the move between keyboard and pointing device. That makes homing one of the largest repeat costs in expert workflows. (textbookofusability.com)
A new metric for your typing tests: Hands‑Off‑Home‑Row (HOHR)
- What it is: the share of total test time spent away from home row due to pointing. You can track it as either (a) % time away, or (b) homing events per minute (H/min).
- Why it helps: HOHR reveals editing efficiency, not just raw WPM.
How to instrument on the web:
- Include mixed tasks (e.g., “Move the caret to line 3, select the next 2 words, wrap in quotes”) that invite pointing. Log pointerdown/mousedown/scroll and blur/focus swaps to estimate hands leaving/returning to keys. Pair that with keystroke logs to segment homing intervals.
- Report: WPM, Accuracy, and HOHR side‑by‑side. Lower HOHR usually indicates fewer context switches and smoother editing.
Predict before you test: use KLM + Homing
KLM lets you estimate time for expert users by summing operators: K (keystrokes), P (pointing), H (homing), M (mental prep), etc. Homing is typically H = 0.40 s. (textbookofusability.com)
- Baseline external mouse or side touchpad: a single “pointing interlude” costs roughly H (leave keys) + P (point) + H (return). If your task needs 30 such interludes, homing alone contributes about 30 × 0.8 s = 24 s.
- Integrated pointing (stick/bar/center touchpad): H shrinks dramatically because your fingers stay on or just off home row. You can model this as H′ ≈ 0–0.2 s per transition depending on device and practice; then recompute.
- Don’t forget P: per‑movement pointing can be slower on isometric sticks or small touch surfaces than on a good mouse. Lab studies show mouse throughput around ~4.9 bits/s vs ~2.9 for touchpads and ~1.8 for joysticks, indicating potentially longer P on integrated devices. So the net win is ΔT ≈ savings from reduced H minus any increase in P. (yorku.ca)
Rule of thumb: If your workflow is “type‑point‑type” heavy (lots of short cursor moves and selections), integrated pointing often wins. If it’s “long drags/precise graphics,” an external mouse still shines.
Validate it: a quick A/B protocol for your site
- Tasks: Build a 5–8 minute “Edit Race” with 40–60 caret moves/select‑and‑fix actions. Include a few precise targets (small links or characters) and a couple of drags.
- Conditions: Test three sessions on different days: (1) external mouse, (2) laptop/desk touchpad, (3) integrated pointer (TrackPoint‑style, roller‑bar, or center touchpad). Randomize order.
- Metrics: Overall time, Accuracy, HOHR, and P‑heavy micro‑splits (time from pointerdown to confirm).
- Analysis: Compare observed gains to your KLM prediction. Many teams see that reduced H offsets a slower P on integrated devices for text‑editing scenarios. Keyboard‑surface pointing research has explicitly argued this time reduction mechanism: keeping pointing on the keyboard shortens mixed typing/pointing tasks. (arxiv.org)
What the research and ergonomics say
- KLM Homing: H ≈ 0.40 s is a widely used constant in practical GOMS/KLM guides and estimators. (textbookofusability.com)
- Pointing throughput: Mouse ≈ 4.9 bps, touchpad ≈ 2.9 bps, joystick ≈ 1.8 bps (ISO 9241‑9 tasks), explaining why per‑movement P is often fastest with a mouse. (yorku.ca)
- Central devices and posture: A 2015 lab study found centrally located touchpads and roller‑mice yielded more neutral shoulder postures, reduced ulnar deviation, and lower forearm extensor activity versus lateral devices — relevant for all‑day editors. (digitalcommons.kettering.edu)
- Integrated pointing reduces travel: Work on “keyboard surface interaction” showed that placing pointing on the keyboard cuts mixed pointing‑and‑typing times by removing the need to reach for a separate device. (arxiv.org)
2026 hardware bringing pointing back onto the keyboard surface
- HHKB Studio (2024): a compact board with a built‑in pointing stick, dedicated mouse buttons, and gesture pads — designed to minimize leaving home row. (hhkeyboard.com)
- Lenovo ThinkPad TrackPoint Keyboard II: a desktop keyboard with the classic TrackPoint nub and buttons, letting you navigate without a separate mouse. (support.lenovo.com)
- Roller‑bar devices (Contour RollerMouse, Mousetrapper): central bars placed in front of the spacebar to reduce reach; studies associate them with more neutral postures and reduced forearm load. (digitalcommons.kettering.edu)
- New in 2026: split mechanical keyboards with a true center touchpad. The Vantex V1, now on preorder, puts a 45 mm touchpad between the halves so your thumbs/fingers can point without leaving the board. (tweaktown.com)
- Also notable: Mokibo’s Fusion line turns most of the keyboard surface into a touchpad — a radical way to keep pointing “on the keys,” though experiences vary by user and firmware. (mokibo.com)
Practical tuning tips for faster edits
- Count your homing: Add HOHR to your results page. If H/min is high, encourage more keyboard shortcuts (Ctrl/Cmd+Arrow, Ctrl/Cmd+Shift+Arrow, Alt/Option+Arrow, Ctrl/Cmd+L, Ctrl/Cmd+K, etc.).
- Tune your pointer:
- TrackPoint/pointing stick: bump sensitivity and press‑to‑select delay; practice micro‑nudges for sub‑pixel control.
- Roller‑bar: position it directly under the spacebar; set low‑to‑medium speed and enable “single‑finger click‑and‑drag” to reduce wrist extension.
- Center touchpad: raise pointer speed slightly above your laptop default; enable tap‑to‑click so thumbs can select without reaching.
- Design your tests: Include both “micro‑moves” (short selections, caret hops) and a few “macro” drags to surface when an external mouse regains the edge.
- Report deltas: Show users “Predicted vs Measured” time saved from reduced homing on their device choice — it’s a great engagement hook.
A quick example
Suppose your 1,000‑word edit has 40 pointing interludes. With an external mouse, homing alone ≈ 40 × 0.8 s = 32 s. Switch to an integrated pointer and assume H′ ≈ 0.2 s each: 40 × 0.4 s = 16 s — a 16‑second saving before any P‑time differences. If your per‑movement pointing is 0.1–0.2 s slower than a mouse, you still come out ahead in text‑centric work. (Use your HOHR logs to calibrate H′ for each device.)
Bottom line
Integrated pointers don’t always move the cursor faster — but they can make you faster overall by slashing homing. Adding HOHR to your typing tests lets you quantify that benefit, predict it with KLM, and help users pick the right device for their real editing work. (textbookofusability.com)
