Skip to main content
. 2022 Nov 22;6(2):e40764. doi: 10.2196/40764

Table 2.

Latent Dirichlet allocation topic modeling for all posts, showing the topics’ post distribution, Facebook metrics, and sentiments from Linguistic Inquiry and Word Count (LIWC; N=34,885).

Topic number Topic name Posts, n (%) Facebook metricsa Sentiments from LIWCb



Number of likes, mean (SD) Number of comments, mean (SD) Number of shares, mean (SD) Word count, mean (SD) Positive emotion percentage, mean (SD) Negative emotion percentage, mean (SD)
1 Heart health promotion 10,912 (31.3) 37.67 (264.95) 2.31 (14.51) 8.81 (50.78) 47.64 (36.70) 5.26 (5.16) 1.12 (2.05)
2 Sharing personal experiences 8094 (23.2) 48.37 (205.55) 8.04 (26.47) 15.14 (113.92) 59.02 (77.90) 4.84 (5.58) 3.02 (4.12)
3 Risk-reduction education 8557 (24.5) 49.63 (217.25) 2.82 (17.36) 18.85 (108.32) 43.04 (57.96) 4.87 (8.56) 4.14 (4.54)
4 Heart disease and heart health promotion for women 5200 (14.9) 68.65 (276.00) 5.56 (24.39) 33.36 (166.07) 44.41 (53.66) 2.53 (3.57) 2.62 (4.03)
5 Educational information sharing 1208 (3.5) 63.65 (114.14) 10.43 (18.39) 5.77 (18.25) 137.98 (40.63) 3.02 (7.25) 2.78 (1.37)
6 Physicians’ live discussion sessions 924 (2.6) 58.21 (152.55) 11.97 (32.09) 14.03 (46.10) 49.2 (34.61) 1.38 (2.10) 3.19 (3.00)

aData collected in November 2021.

bPositive and negative emotions represent the percentage of words in a post that appear in the dictionary indicating positive and negative emotions.