Skip to main content

Table 2 Results of maximum likelihood estimation on ERGM parameters

From: Network analysis of gratitude messages in the learning community

Term

Dependent variable: Estimate

Parameter

Node-based

Dyadic

Structural

edges

Messages

 − 6.475*** (0.889)

 − 5.097*** (0.222)

 − 4.229*** (0.110)

nodecov.score

Evaluation score by an instructor

0.056 (0.046)

  

nodecov.character

Number of characters in messages

 − 0.0001 (0.0001)

  

nodecov.access

Number of discussion forum access iterations

0.005* (0.003)

  

nodecov.gratitude

Number of messages with gratitude

0.218*** (0.053)

  

nodecov.EGRM

Score of questionnaire

0.020 (0.030)

  

nodecov.link

Number of web links in messages

 − 0.003 (0.031)

  

nodecov.fig

Number of figures in messages

0.070 (0.063)

  

nodefactor.gender.female

Female students

 − 0.109 (0.151)

  

mutual

Reciprocal messages

 

3.184*** (0.298)

 

nodematch.country

Same country

 

0.994*** (0.267)

 

nodematch.question

Same question number

 

0.926*** (0.142)

 

nodematch.gender

Same gender

 

0.376** (0.170)

 

absdiff.score

Homophily in evaluation score

 

 − 0.114** (0.060)

 

absdiff.character

Homophily in character count

 

 − 0.0001 (0.0001)

 

absdiff.link

Homophily in number of web links

 

0.044 (0.031)

 

absdiff.fig

Homophily in number of figures

 

 − 0.088 (0.075)

 

absdiff.access

Homophily in number of discussion forum access iterations

 

0.003 (0.003)

 

absdiff.gratitude

Homophily in number of gratitude messages

 

0.129** (0.065)

 

absdiff.EGRM

Homophily in score of questionnaire

 

 − 0.059 (0.038)

 

istar(3)

Incoming 3 edges

  

 − 0.046 (0.072)

gwesp.ISP

Incoming shared partner

  

 − 0.535 (0.677)

gwesp.ITP

Multiple cyclic closure

  

0.335* (0.201)

  1. Upper, parameter estimate value; lower, parameter standard error
  2. *p < 0.1; **p < 0.05; ***p < 0.01