In-store music and wine sales
P(X ≥ 25)= 1-P(24 ≥ X)= 0.00037
As a result, from the above hypothesis test, it can be concluded that we can accept the alternative hypothesis since 0.00037<0.01 and that the proportion of French wine bottles bought has increased significantly. Thus the addition of French music has had a significant impact in raising the proportion of French wine sold, as the majority of all external factors were controlled. Furthermore, the data collected from the survey asking ‘What is the country of origin of the music that is being played?’ suggested that the music used was obviously French to the consumers and thus could trigger a response to the stimuli. This was apparent as all 40 of the 40 randomly selected customers successfully identified the music as French. However, although the data collected suggests a resounding success on the French side of the experiment, there were some limitations to the data collection process that, to some extent, invalidate the result. First, the experiment with French music was only carried out over one day with no repeats. As a result, the reliability of the data can be brought into question as it is unclear whether this result is consistent. Similarly, the reliability of the data can be questioned given some of the flaws in the design of the experiment, particularly the day of the week and time of the year the data was taken from, compared with the control data from previous sales reports. Firstly, the data in the presence of French music was collected on a Saturday, the shop ’ s busiest day on average, while the control data takes all days of the week into account. Consequently, there may be a major difference in the data collected on a Saturday against data from every other day of the week which could have a skew effect on the data and cause an overestimation in the percentage increase in French wine bought. To improve the reliability of the data, it should have been compared to store data from all the Saturdays in the last month, thus rendering a more accurate comparison to base the hypothesis test on. However, this data isn’t collected by the shop , so it wasn’t available to me. Finally, the extent to which priming was the sole cause of this increase in French bottles being bought may have been limited by the impact of anchoring bias. As can be seen in figure 3, the first wines seen when entering the shop are French wines which may have caused a biased response due to the anchoring effect where customers may have been drawn to the wines they saw first. As a result, it is more difficult to measure the direct correlation between playing French music and the increase in the number of French wine bottles bought as a complete side effect of priming. On the final day of the experiment, only Italian music was played in the shop and the procedure was kept identical to the previous 2 days. According to figures 1 and 2, Italian wine underperformed in sales when compared to Italy’s status as a world renowned wine producer. Over the previous one month only 4 bottles were bought daily (on average), only 11% of the total share. Therefore what was being examined in this part of the experiment was quite different to that with French wine: I was testing whether priming could cause a poorly performing wine to excel in sales despite the clear consumer Figure 3: A photo from the doorway of the wine shop ‘Vinomondo’: with French wine being on display straight ahead
188
Made with FlippingBook - PDF hosting