Strength of Weak Ties (Granovetter)
Strength of Weak Ties
(Granovetter)
What is a weak tie?
• Strength of tie as a linear combination• F = Frequency of Contact• E = Emotional intensity• I = Intimacy (mutual confiding)• R = Reciprocal services
lij=w1F+w2E+ w3I+ w4R
• Granovetter leaves exact functional form and weighting “postponed for future empirical work”
Simplification
• For purpose of building theory, ties classified as • “strong”• “weak”• “absent”
Example
A B
S={C,D,E…}
Connection Closure
• If • A is connected to B (strong tie)• A is connected to C (strong tie)• A-B and A-C are independent
• Then• P(B-C | A-B, A-C) = P(A-B)*P(A-C)
• A weaker connection B-C exists with a higher probability
• Common strong ties generate new connections
Cognitive Balance
• Anything short of a positive tie between B and C “would induce psychological strain”
The “Forbidden Triad”• The amount of dissonance
between B and C is proportional to strength of ties A-B and A-C
• If A-B and A-C are strong, triad occurs rarely
• If A-B and A-C are weak, dissonance is low and weak tie B-C can be present
A B
CDissonance
Connections and Similarity
• Newcomb(1961), Friedkin (later)• The stronger the tie connecting two
individuals, the more similar these individuals are
• Thus, if A-B and A-C are strong ties• B is similar to A, C is similar to A
• By transitivity
• B is similar to C
Weak ties and Bridging
• Strong ties are unlikely to be bridges
• Weak ties can bridge distinct groups without interpersonal dissonance
QuickTime™ and aTIFF (LZW) decompressor
are needed to see this picture.
Local Bridges
• Global bridges (I.e. cutpoints) are unlikely in real networks• Single ties between “clumps” are unlikely
• Local bridges create paths between groups• Local bridge of degree n
• Weak ties allow for more redundancy• Shorter paths
Diffusion
• Weak ties speed up information diffusion• Shorter distances mean faster penetration• More redundancy means wider reach
Weak Ties in Ego Nets
• Burt: look at overall density and redundancy in ego nets (I.e. Structural holes)
• Granovetter: Partition the ego network into three groups:• Strong ties• Weak ties
• Personal ties of the Ego
• Bridging weak ties• Guaranteed to connect ego to outside groups
Empirical Study
• Job search in Boston• Survey asks how applicant found out about
the job opening• If job is found through personal contacts:
• How often does ego see the contact?• Where did the contact get the information?
Hypothesis:
• These with strong ties to ego are more motivated to help find a job
• Weak ties connect to information outside immediate group
• Information gets diffused by long paths
0
10
20
30
40
50
60
0 200 400 600 800 1000 1200
Result
• 16.7% report strong ties (interact often)• 55.6% report mid-strength
(occasionally)• 27.8% interact rarely
• Often > twice a week• Occasionally > once a year• Rarely < once a year
Questions
• Are these gradations sufficient?• Do we need greater granularity?
Questions
• What is possible distribution of overall tie strength in the network? • Skewed normal?• Maybe a power law?
• We know that degree distributions tend to power laws
• Does it affect what ties have higher probability of being used for job search?
More results
• Source of job information:• 39.1%: direct from employer• 45.3%: one intermediary• 12.5%: two• 3.1%: > 2
0
5
10
15
20
25
30
35
40
45
50
0 0.5 1 1.5 2 2.5 3 3.5
Leadership and weak ties
• High number of strong ties in the in-group generally means less ties to the out-group• Related to limited cognitive capacity?
• Societies characterized by tight in-groups emerge local leaders, but each in-group acts on its own
Weak ties in politics
• Trust in a leader a function of having a bridging tie from ego to leader • (even a very weak one)
Weak ties in politics
• Politicians must cultivate weak ties near election time• “kissing babies”• Name recognition
• Strong ties are required to get things done• Committee work