There were limitations to the predictive power of Google Trends when it came to elections. There was no evidence of a strong correlation between search popularity and the likelihood of winning. congressional elections concluded that, compared to the traditional methods of election forecasting, Google Trends did not prove to be a good predictor of either the 2008 or 2010 elections. įurthermore, a 2011 Wellesley College study of the 20 U.S. ICONIC claimed that Google data enabled the model to bypass flaws that were inherent in traditional polls and surveys such as non-response bias, voter shyness and difficulties in identifying the correct voter proportions. In fact, online jewelry retailer ICONIC used Google Trends data to predict a Trump victory with 93% probability. Trump had a wide lead over Biden in Google Trends even as polls predicted a Biden victory. election did not have the same predictive power as in previous elections. However, this would not be the case in 2020. In a specific case, search interest in Trump actually correctly predicted the election result in 2016. The results showed that Google Trends had predicted the real winner in these elections held since 2004, but the paper went only as far as recommending to monitor the November 2020 presidential elections in the United States and to have more accurate information on future results. Another paper published in November 2020 analyzed the predictive capacity of Google Trends in four presidential elections in the United States and the five presidential elections in Canada beginning 2004. These preferences however were influenced by various social events which should be filtered out as noise for better estimation of the results. Specific case studies on the elections in Germany purportedly showed strong correlation between the Google search preferences of users and the actual election results. Hence the main question, “Can Google Trends volume predict the outcome of elections, or is it merely coincidence?” We know that there is a big difference between the demographics of voters and the people who use web tools like Google. This is taken from all web searches in the recent week in the Philippines on all categories. For example, if we use the surnames only of the top five candidates in the coming Philippine presidential elections, we get as of this writing the following rather different result contrary to what others have shown: Marcos (53), Robredo (23), Pacquiao (5), Moreno (6) and Lacson (4). Google Trends is also very specific on the terms or topics being searched for comparison. Using the graph, it is quite tricky and oftentimes misleading when we compare search trends between multiple keywords. Thus, a downward trend on the graph means that a search term’s relative popularity is decreasing, but it does not mean that the total number of searches for that term is decreasing. It takes a small sample of the absolute search volume then normalizes the data on a scale from 0 to 100 where each point is divided by the highest point on the graph. Google Trends is a web tool that gives numeric and visual representations of the popularity of a term or topic searched on Google over a specific period and location relative only to the total number of searches.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |