Twitter psychometry predicts DJIA for 2-6 days
Johan Bollen (School of Informatics at the University of Indiana in Bloomington) and a group of his colleagues still published a sensational work on the statistical correlation of attitudes on Twitter and the Dow Jones index (DJIA). They made an announcement of their research earlier and the named figures on the accuracy of coincidence surprised their colleagues a lot. On October 14, 2010, their entire scientific work was published ( PDF ).
Johan Bollen and colleagues used Granger’s causality test to analyze the time series of the Dow Johns index and Twitter sentiment for the period from February 28, 2008 to November 3, 2008 on weekdays. To identify emotions in tweets, OpinionFinder and GPOMS linguistic analysis system was used - the Profile of Mood States psychometric system (POMS-bi), expanded with 2006 and 4-letter words compiled by Google .
The researchers tried different variants with a shift in time and found that in one of the parameters the accuracy of the matches leaves no room for randomness.
The table shows the statistical significance (p-values) of the two-dimensional correlation in the Granger causality test between the moods and the DJIA from February 28, 2008 to November 3, 2008.
')
As stated in the scientific work, “based on the values ​​of the Granger causality test, we can reject the null hypothesis with a high degree of confidence,” that is, the hypothesis that there is no relationship between the two data sets. One of the parameters of psychometry (calm, calm) shows a very high correlation with the DJIA with a shift of 2-6 days, and here the results can be called statistically significant.
The “calm” value predicts the movement of the index with a probability of 87.6%. If you put them on one graph, then there are noticeable gaps when the mood on Twitter does not affect the index at all. According to scientists, these days unpredictable factors influence the market. For example, an unexpected statement by the Fed to change the accounting rate or quantitative easing, etc.
The research methodology is shown in the diagram.
Johan Bollen and colleagues used Granger’s causality test to analyze the time series of the Dow Johns index and Twitter sentiment for the period from February 28, 2008 to November 3, 2008 on weekdays. To identify emotions in tweets, OpinionFinder and GPOMS linguistic analysis system was used - the Profile of Mood States psychometric system (POMS-bi), expanded with 2006 and 4-letter words compiled by Google .
The researchers tried different variants with a shift in time and found that in one of the parameters the accuracy of the matches leaves no room for randomness.
The table shows the statistical significance (p-values) of the two-dimensional correlation in the Granger causality test between the moods and the DJIA from February 28, 2008 to November 3, 2008.
')
As stated in the scientific work, “based on the values ​​of the Granger causality test, we can reject the null hypothesis with a high degree of confidence,” that is, the hypothesis that there is no relationship between the two data sets. One of the parameters of psychometry (calm, calm) shows a very high correlation with the DJIA with a shift of 2-6 days, and here the results can be called statistically significant.
The “calm” value predicts the movement of the index with a probability of 87.6%. If you put them on one graph, then there are noticeable gaps when the mood on Twitter does not affect the index at all. According to scientists, these days unpredictable factors influence the market. For example, an unexpected statement by the Fed to change the accounting rate or quantitative easing, etc.
The research methodology is shown in the diagram.
Source: https://habr.com/ru/post/106368/
All Articles