ITR Atlas
← All interfaces

Stylus/Tablet Pointing (RAND, 1964)

Measured by MacKenzie, Sellen & Buxton · ACM CHI 1991 (1991)

Stylus Pointing 1964

Inputs

The measured or assumed values behind the calculations, each with its source.

TP = 4.9 bits/s
Fitts' law throughput for the stylus/tablet in the 1991 pointing task, the highest of the three devices tested (mouse 4.5, stylus 4.9, trackball 3.3). The system date is the 1964 RAND Tablet / GRAIL graphical input lineage.

Strictest ITR

Each scoring method is an upper bound on the channel, so the headline is the strictest (smallest) one for this entry. Use the score selector on the home page to view any single method across entries.

Strictest Fitts' law Author-reported · reproduced
Fitts' law throughput, re-derived from the task conditions
2D pointing channel
4.9 bits/s

  1. Information per movement (index of difficulty)

    ID = log2(A/W + 1);  4 amplitudes (8–64) × 4 widths (1–8) fully crossed → ID = 1.0–6.0 bits (mean 3.26 bits/movement)

    Each pen-on-tablet movement selects among the distinguishable endpoints set by the distance-to-width ratio; that ratio, in bits, is the Shannon information per movement. Task design: MacKenzie, Sellen & Buxton 1991.

  2. Accuracy folded in via effective width

    W → We for the observed 4.0% error rate (Welford normalization, 4% nominal)

    Effective width discounts movements whose endpoints scattered past the target, so throughput is net of the user's real precision.

  3. Throughput = information ÷ movement time

    mean MT = 665 ms → IP = 4.9 bits/s (Fitts regression; edged out the mouse for pointing)

What counts as a bit depends on the action space. The number of distinguishable actions and how likely each one is are design choices of the task, not the sensing hardware. The same modality can present a fixed set of targets, a set pruned per step by a grammar or language model, or a continuous control space. Each of these changes how many actions are live and how the probability mass is spread, and therefore the information per selection. Read the action space below before comparing headline numbers across entries.

Action space

What the user can produce at each step, and how those options are distributed.

Structure
Continuous control space
Size
Continuous
Prior
Uniform: all actions assumed equally likely
Notes
Direct pen-on-tablet pointing. Throughput edged out the mouse for pointing in this study (though not for dragging). Continuous control channel, comparable to the cursor BCIs' bits/s, not to discrete spellers.

Comparability The strictest bound here is Fitts throughput: the index of difficulty, log₂(A/W + 1), per movement. Directly comparable to the other continuous-pointing entries (mouse, trackball, stylus, gaze and the cursor BCIs). Set against the text entries (keyboards, spellers, speech) it crosses methods: both report bits/s, but one measures movement difficulty and the other text information, so compare within the family first.

Source

Authors
MacKenzie, Sellen & Buxton
Publication
ACM CHI 1991, 1991
Paper
10.1145/108844.108868
Reference
System date: RAND Tablet / GRAIL stylus system