Social shapes challenge: method

What it measures

Social intelligence / mentalizing: inferring intentions, relations and emotions from the motion of simple geometric shapes. The clips contain no faces, voices or words; the questions and answer options are text.

Procedure

14 silent animations (about 10 to 23 s each) shown in a fixed order. Each clip must play to the end before an answer is accepted. Then one multiple-choice question with 4 options. Replay is allowed, and the player can go back to change an answer. No time limit, no practice item, no feedback during the run. Input: tap or click, or the keyboard (Tab and arrow keys, then Next). All 14 questions must be answered for a result.

Parameters

Scoring

Number of answers that match the original key, 0 to 14; higher is better. No subscores and no scaling to an IQ-like metric. Compared with other players only as a band (below typical = under the 25th percentile, typical, above typical = over the 75th) once 100 comparable players exist; before that the page shows how many have played. Player comparisons use only first runs (no earlier run by the same visitor, account, or IP address and device) without quality flags, with the same scorer version and the same runner settings.

Valid-run rules

Every item needs exactly one answer, and the item-bank version must match the current one (otherwise the page must reload). An answer is rejected unless its clip was marked as watched and the time on the item was at least the clip length (with 250 ms tolerance for video container differences). Quality flag 'too fast': at least half of the answers given within 700 ms of the item appearing. The server rebuilds every run from the raw log (and the run's seed where there is one) and computes the score itself; anything the page says about right or wrong is ignored. A log that breaks the game's rules is rejected, and nothing is stored. Stored runs can carry quality flags that are never shown and never block a result: 10 or more runs from one IP address, a missing visitor id, tab switches, and automation signals (automated browser, impossible speed, no input events, machine-regular timing).

Differences from the classic task

The published Social Shapes Test uses fixed forms (a 23-item final set and two 14-item parallel forms). This page uses its own selection of 14 scored items from the final 23-item set, which matches neither parallel form. Replay and going back are allowed. Published reliability, validity data and norms do not transfer to this selection or to the translations; results are compared only with other players of this page.

Limitations

Unsupervised and untimed. Reading skill and the language of the questions affect the result, and the translations have not been validated. Knowing the stories makes a second run easier. With 14 items and 4 options, one run is a coarse measure. It cannot diagnose autism or any other condition, and it does not describe how a person behaves in real conversations. The comparison players are site visitors, not a representative sample.

Languages and review status

The animations are silent and the same in every language. Question and option text: English original plus 99 other languages (machine-assisted translation with automated script and spelling checks, comparison with native text corpora). Translations keep the original answer key and option order.