Stats › Figures exploitables › Social media signals
The famous Twitter mood result rests on fifteen test days and does not match its own table
| Mood series | Lag 2 | Lag 3 | Lag 4 | Lag 5 | Lag 6 |
|---|---|---|---|---|---|
| Calm | 0.013 | 0.022 | 0.030 | 0.036 | 0.065 |
| General positive and negative | 0.268 | 0.436 | 0.218 | 0.300 | 0.446 |
| Alert | 0.973 | 0.981 | 0.998 | 0.989 | 0.996 |
| Sure | 0.811 | 0.349 | 0.415 | 0.544 | 0.691 |
| Vital | 0.369 | 0.418 | 0.475 | 0.553 | 0.682 |
| Type | 0.991 | 0.991 | 0.989 | 0.996 | 0.994 |
| Happy | 0.706 | 0.723 | 0.750 | 0.173 | 0.081 |
Granger causality p values from the original paper, seven mood series against seven lags, 9,853,498 tweets from about 2.7 million users between 28 February and 19 December 2008, with 342,255 mood bearing tweets in the causality window. One series of seven is significant at five per cent across four of seven lags and no correction for multiple testing is reported anywhere. The paper's text says the significant series holds at lags two to six, while its own table shows lag six at 0.065.
The 2011 paper claiming social media mood predicts the market reports 87.6 per cent directional accuracy in its abstract and 86.7 in its own table, over fifteen trading days, with 49 untested hypotheses behind it.
Réservé aux membres
La suite de cette page est réservée aux membres
Below this point there are 4 sections and 3 named sources, roughly 800 words of it. Every figure carries the source it came from and the date the data is from.
Vous pouvez continuer à parcourir gratuitement toutes les statistiques de la bibliothèque. L'introduction et le résumé sont toujours accessibles.