Case study to show variations in Perceived Climate Index of travelers in conjunction with the weather condition at a destination.
There is a lot of data being generated on social media constantly and many a times, the insights drawn from it are questioned to examine the reliability of the same. In this newsletter, we present to you a case study that shows how real time changes in local weather conditions at a destination, influences the perception of travelers on social media. This perception is measured using Socialvane’s Perceived Climate Index (PCindex).
With the summer already begun in Spain, tourists started flowing to Palma de Mallorca in June to enjoy the warm sun on the beach. In fact, in June last year, Mallorca received more than 1.39 million travelers from domestic and international markets (source). These numbers reinforce the fact that sunbathing is a popular touristic offering in Mallorca.
Hence, we chose this destination to show the correlation between changes in weather conditions using temperature measurements in degree centigrade and our PCindex, which measures the degree of satisfaction with the climate for travelers at the destination. To do so, we monitored daily measurements of both, temperature and our PCindex in Palma de Mallorca during the month of June 2016.
While the weather was mostly sunny from the beginning of June, there was an unexpected spell of rain during the 3rd week of last month. We can see from the above graph that the blue line representing the PCindex which has a score of around 90 points on regular sunny days sees a steep fall in the score the days it rained, which means there is a direct effect on the PCindex of the destination. Also, during the 4th week of June, the steep decrease of our PCindex coincides with the fall in temperatures in Palma.
We calculated the correlation between the daily changes in temperature and the PCindex by using Pearson coefficient and we got an average value of 0.38, implying a moderate relationship.
Even though the average correlation between the two variables is a moderate 0.38, the correlation among the two is strong on the days when it rained. We can see that, when it rained on two different days on the 18th and 29th of June, the pattern in the fall in PCindex is perfectly the same. Suggesting that higher variations in the weather condition have a direct, linear impact on our PCindex.
Even though there is no progressive relationship degree by degree, there is an evident fall in the PCindex when the temperature falls for a sunny destination like Palma de Mallorca during summer. Which is a nice way to see how social data can be related to “real world” measurements such as temperature or rain.
Of course, the PCindex measures travelers’ perceptions and expectations about a destination. This means that we will find different behavioral patterns in different destinations. This is precisely what makes this data relevant and interesting: the fact that we are now able to have a better understanding of how specific events affect travel dynamics in each destination specifically, using social data.