Skip to main content

Table 2 Goodness of fit

From: More is not always better—non-linear effects in crowdfunding

Criterion

b-spline

tp-spline

Linear regression

R2 updates

.21

.21

.17

R2 comments

.19

.20

.11

R2 rewards

.07

.07

.06

R2 Facebook friends

.00

.01

.00

R2 Facebook shares

.03

.03

.02

R2 Twitter tweets

.01

.01

.01

Aggregated R2 control variables

.45

.45

.45

R2 overall

.54

.55

.51

BIC

96,097.40

95,247.54

104,611.60

Log-likelihood

− 47,729.23

− 47,154.90

− 52,056.04

d.f.

110,918

110,972

110,930

  1. R2 Variance explained adjusted for degrees of freedom for different model effects
  2. BIC Bayesian information criterion (lower is better), Log-likelihood value of probability function for model (lower is better), d.f. degrees of freedom for model
Back to article page