In this work we test if Internet access device impacts the online purchase probability of a consumer. We also test if the online purchase probability of consumers increases over time. Our data is from the Canadian Internet Use Survey over four years, consisting of at least 23,178 participants per year, relative to previous studies this is a very large dataset. We segment the participants into two groups: 1) those that use only their personal computer (PC) to access the Internet and 2) those that use their PC and a mobile device (mobile phone, tablet, etc.) to access the Internet. For each group we determine the purchase probabilities of consumers for 15 different product category types. Using 95% confidence intervals about the determined purchase probabilities we test if the purchase probability of one group is higher than the other. We find that mobile devices lead to a statistically greater purchase probability in only 4 of the 15 product categories. In addition, the purchase probability of households using mobile devices and PCs to access the Internet only increases between years 2005 and 2007, and is statistically unchanged between all other years. However, for households that use only a PC to access the Internet, the purchase probability increases each year the survey was conducted.