Table 3 Derived metrics for domains.

From: DomainDemo: a dataset of domain-sharing activities among different demographic groups on Twitter

Metric name

Size

Range

Data universe

Calculation

Note

Localness

129,127

[0, + )

state

Eq. (2)

Larger values indicate more local domains.

Race deviation

129,127

[0, + )

race

Eq. (2)

Larger values indicate that the domain shares are more concentrated on some race categories.

Age deviation

129,127

[0, + )

age

Eq. (2)

Larger values indicate that the domain shares are more concentrated on some age categories.

Audience partisanship

129,127

[ − 1, + 1]

party

Eq. (3)

Only including Democrat and Republican categories. Negative values indicate more shares from Democratic users and vice versa.

Audience partisanship (registration)

129,041

[ − 1, + 1]

party registration

Eq. (3)

Only including Democrat and Republican categories. Negative values indicate more shares from Democratic users and vice versa.

Gender leaning

129,127

[ − 1, + 1]

gender

Eq. (3)

Only including Male and Female categories. Negative values indicate more shares from Male users and vice versa.

  1. The table presents six metrics derived from six different data universes. From left to right, we report the metric name, size, range, data universe, calculation method, and a brief description.