Knowledge Base

Frequency/Sequence and Gaps; Mutually Exclusive, Non-Exhaustive Behavior Groups

Example

The figure below shows the visualization of two observations, Obs 1 and Obs 2, with events (A, B) from a group in which behaviors exclude each other. The behaviors in the group are non-exhaustive, which means that gaps between events may occur.

Obs 1 contains two gaps:

• From 1.3 to 3.2 s.
• From 12.3 s till the end of the observation.

Without Analyzing Gaps

When you de-select Analyze gaps between events, two agreements and one disagreement are found.

• Agreement: 0.0 A (Obs 1) - 0.0 A (Obs 2).
• Disagreement: 3.2 A (Obs 1) - 0.0 B (Obs 2).
• Agreement: 7.9 A (Obs 1) - 7.9 B (Obs 2).

With Analyzing Gaps

When you select Analyze gaps between events, the comparison of observations 1 and 2 gives two extra disagreements between a gap in observation 1 and an event in observation 2.

This results in the following:

• Agreement: 0.0 A (Obs 1) - 0.0 A (Obs 2).
• Disagreement: 1.3 <Gap Group AB> (Obs 1) - 0.0 A (Obs 2).
• Disagreement: 3.2 A (Obs 1) - 0.0 A (Obs 2) (this is a disagreement because 0.0 A in Obs 2 was already in agreement with event A in Obs 2).
• Agreement: 7.9 B (Obs 1) - 7.9 B (Obs 2).
• Disagreement: 12.3 <Gap Group AB> (Obs 1) - 7.9 B (Obs 2).

If during an observation no behaviors from a mutually exclusive non-exhaustive group are scored, Reliability Analysis will consider a gap for that group from the start to the end of the observation.

Duration/Sequence and Gaps; Mutually Exclusive, Non-Exhaustive Behavior Groups

Example

The figure below shows the visualization of two observations, Obs 1 and Obs 2, with events (A, B) from a group in which behaviors exclude each other. The behaviors in the group are non-exhaustive, which means that gaps between events may occur.

Obs 1 contains two gaps:

• From 1.3 to 3.2 s.
• From 12.3 s till the end of the observation (16.0 s).

Obs 2 contains one gap:

• From 14.9 s till the end of the observation (16.0 s).

Without Analyzing Gaps

When you de-select Analyze gaps between events, the comparison of Obs 1 and Obs 2 results in the following three agreements and two disagreements:

• Agreement: from 0.0 till 1.3 s; Event A overlaps; duration 1.3 s.
• Disagreement: from 1.3 till 3.2 s; gap in Obs 1 and Event A in Obs 2; duration 1.9 s.
• Agreement: from 3.2 till 7.9 s; Event A overlaps in both observations; duration 4.7 s.
• Agreement: from 7.9 till 12.3 s; Event B overlaps in both observations; duration 4.4 s.
• Disagreement: from 12.3 till 14.9 s; gap in Obs 1 and Event B in Obs 2; duration 2.6 s.

With Analyzing Gaps

When you select Analyze gaps between events, this example contains an extra agreement, because there is an overlapping gap of 1.1 s at the end of both observations.

Frequency/Sequence and Gaps: Start-Stop Behavior Groups

Gaps in Start-Stop Groups

In a behavior group in which behaviors may overlap, each behavior is analyzed independent from the others. The frequency/sequence method searches for matches between event A and event A, and between event B and event B (see Start-Stop groups for a detailed explanation). When you analyze gaps between events, the same is true for the gaps in such groups. For each individual behavior is recorded whether there is a gap or no gap. So the program compares Gap A with Gap A, and Gap B with Gap B.

Note: If during an observation a start-stop behavior is not scored, Reliability Analysis will consider a gap for that behavior from the start to the end of the observation.

Example

Without Analyzing Gaps

When you do not analyze gaps between events, the Reliability Analysis gives the following two run 1 agreements, because the events partially overlap in both observations:

• Agreement: 0.0 A (Obs 1) - 3.2 A (Obs 2)
• Agreement: 9.5 B (Obs 1) - 5.5 B (Obs 2)

With Analyzing Gaps

When you select Analyze gaps between events, two agreements and one disagreement are added, so you obtain four agreements and one disagreement:

• Agreement: 0.0 Gap B (Obs 1) - 0.0 Gap B (Obs 2)
• Agreement: 0.0 A (Obs 1) - 3.2 A (Obs 2)
• Disagreement: 0.0 A (Obs 1) - 0.0 Gap A (Obs 2)
• Agreement: 5.5 Gap A (Obs 1) - 6.1 Gap A (Obs 2)
• Agreement: 9.5 B (Obs 1) - 5.5 B (Obs 2)

Duration/Sequence and Gaps: Start-Stop Behavior Groups

Gaps in Start-Stop Groups

In a behavior group in which behaviors may overlap, each behavior is analyzed independent from the others. The duration/sequence method searches for overlaps between event A and event A, and between event B and event B (see Start-Stop groups for an explanation). When you select Analyze gaps between events, the same is true for the gaps in such groups. For each individual behavior is recorded whether there is a gap or no gap. So the program compares overlaps of Gap A with Gap A, and Gap B with Gap B.

Example

Without Analyzing Gaps

When you deselect Analyze gaps between events, comparison of observations 1 and 2 results in the following agreements and disagreements:

• Disagreement: from 0.0 till 3.7 s; Event A in Obs 1 and Gap A in Obs 2; duration 3.7 s.
• Agreement: from 3.7 till 5.5 s; Event A overlaps; duration 1.8 s.
• Disagreement: from 5.5 till 6.1 s; Gap A in Obs 1 and Event A in Obs 2; duration 0.6 s.
• Disagreement: from 5.5 till 9.5 s; Gap B in Obs 1 and Event B in Obs 2; duration 4.0 s.
• Agreement: from 9.5 till 16 s; Event B overlaps; duration 6.5 s.

With Analyzing Gaps

When you select Analyze gaps between events, extra agreements are found for Gap B from 0.0 till 5.5 s and for Gap A from 6.1 till 16 s = 9.9 s.