LogosLink User's Manual · LogosLink version 2.0.0

Participation by Unit Analytics (Argumentation Model)

Participation by Unit analytics shows how much each speaker in an argumentation model participates in the conversation.

This is useful to quickly visualise who speaks more and who speaks less in a debate or discussion.

Parameters

Unit. This indicates what unit to use to segment the text into blocks. You can choose between locutions, words, and word-forming characters.
Block length. When run, this analytics segments the text into blocks of equal length for the selected unit. This parameter establishes the length of this block. For example, if you selected "Words" as unit and "20" for block length, the text will be segmented in blocks of 20 words each. Typical values are between 5 and 20 locutions, 50 and 200 words, or 500 and 2000 word-forming characters.

Results

Results are given as a series of charts. All the charts show a pseudo-temporal horizontal axis that spans the selected units in the argumentation model, divided in blocks of the selected length.

Locutions

The chart shows each locution over time as a colour block, using different colours for different speakers.

This chart is useful to quickly see how much each speaker actually speaks, and when they do it.

Counts

The chart shows how many speakers participate in each block, and and with how many locutions.

This chart is useful to quickly see how much was spoken for each block, and by whom.

Units

The chart shows how many units each speaker spoke during each block.

This chart is useful to quickly find which speakers dominate the conversation at each moment.

Utterance Speed

The chart shows how "fast" each speaker spoke during each block, in word-forming characters per unit. An average utterance speed for all speakers is also shown.

This chart is useful to quickly find who speaks faster and slower, and above or below the average.

Details

This analytics can replace the Participation by Time analytics when the locutions in the argumentaton model lack start and end time information.

By segmenting the text in blocks of a fixed unit size, and using the selected unit as a proxy for actual time, the analytics can provide results that might be close to those provided if time information had been available.