Table 1 The ∆AICc for models explaining variation in the: (1) mean speed of individuals, (2) mean group cohesion, (3) group exploration (the total number of moves individuals made into arms), (4) proportion of moves into arms that were leadership attempts, (5) number of leadership attempts, and (6) number of following events.

From: Group personality, rather than acoustic noise, causes variation in group decision-making in guppy shoals

 

(1)

  

(2)

  

Response variable

Mean speed

  

Group cohesion

  
  

∆AICc

d.f.

 

∆AICc

d.f.

Explanatory variables

Treatment + day + start time

0

6

Start time only (null model)

0

4

Day + start time

0.3

5

Treatment + start time

0.5

5

 

Treatment * day + start time

2.4

7

Day + start time

0.7

5

 

Treatment + start time

5.3

5

Treatment + day + start time

1.1

6

 

Start time only (null model)

5.7

4

Treatment * day + start time

3.3

7

 

(3)

  

(4)

  

Response variable

Group exploration

  

Proportion of leadership attempts

 
  

∆AICc

d.f.

 

∆AICc

d.f.

Explanatory variables

Treatment + day + start time

0

6

Treatment + start time

0

4

Treatment * day + start time

0.3

7

Treatment + day + start time

0.3

5

 

Day + start time

0.5

5

Start time only (null model)

1.9

3

 

Start time (null model)

5.1

4

Day + start time

2.2

4

 

Treatment + start time

5.3

5

Treatment * day + start time

2.7

6

 

(5)

  

(6)

  

Response variable

Number of leadership attempts

  

Number of following events

  
  

∆AICc

d.f.

 

∆AICc

d.f.

Explanatory variables

Day + start time

0

5

Treatment + day + start time

0

6

Start time only (null model)

1.3

4

Day + start time

1.5

5

 

Treatment + day + start time

2.2

6

Treatment * day + start time

1.9

7

 

Treatment * day + start time

3.5

7

Treatment + start time

7.4

5

 

Treatment + start time

3.5

5

Start time only (null model)

7.8

4

  1. The null model includes start time as the only fixed effect. All models include group identity as the random effect. Treatment is noise or control, day is the day of testing (1 to 4).