Main Page
Pew Data Analysis
OkCupid Data Analysis
Final Thoughts
Future Directions
Presentation Materials

With information from both data sets, we tried to find interesting ways to visually compare them.

First off, we took a second look at the chord diagram to see what it could tell us. Remember: one chord represents one data point. The more area the chords take up, the more intersection there is between two demographics.

This diagram shows that both the majority of online daters and average people are composed of whites. This makes sense because the surveys we used were collected in US, where whites are the dominant population. We can see that there are not many black people who use dating site (specifically OkCupid). This doesn't mean that black people are not interested in using other dating sites,It might just mean that most OkCupid users favor non-black people, which drives black people away from using OkCupid. The other interesting thing we see is that there are more asian online daters than the average asian people. We didn't expect that there would be over 50% of asian population, from our datasets, on dating sites, given their cultural background. But, again, this data might not represent the whole asian population in US as they are mostly from San Francisco Bay Area, which have more open-minded and modernized asians who do not let themselves constrained by their traditions. We can also see that there are only a few native americans on dating sites. It may be because native americans tend to prefer meeting someone of their own race in a traditional way rather than meeting someone who they know nothing about online.

Having analyzed the chord diagram, we did some other side-by-side comparisons of the data.
Below is the age distribution of each data set. (Click to view larger)

The histogram on the left shows how most of the users of OkCupid typically are around the range of 20-35 years old, which is most likely due to people wanting to get married in these years. The low amount of people with ages under 20 is probably because OkCupid does not accept users under 18. After age 35, you can see that the graph begins to taper off. This is most likely because people that age have found matches, or they believe in traditional dating and don't think that online dating is the right way to meet the right match for them.

The histogram on the right from the Pew Data shows a much different age range than the OKCupid age distribution did. The average age shown here is around 50-70, with the most common age lying around 60 years old. There are also quite a few younger respondents around the ages of 20-30 again. The more even distribution in this graph could be due to online daters and traditional daters being grouped together in this graph.

We also looked at the income distribution reported from each data set.

In the OkCupid histogram on the right, we can see that there were a lot of users with a more average income, but there seems to be almost a hole in the graph before income shoots back up again. This could be that 80% of people didn't want to report their income because they believed it to be too low to attract a significant other, or they didn't want money to be a factor in their relationship. There are also quite a few people in the higher income range.

This nationwide graph shows a more likely distribution of income than the OKCupid graph probably because more people are willing to vocalize what they income is when they don't have a relationship on the line. As you can see there is a more steady picture of the income, and still quite a few wealthy individuals.

We also attempted to compare the sentiments gathered from OkCupid to the satisfaction reported in the Pew Data.

By the graph on the left we can tell that most of the sentiments had no value, and if they did have value it seemed to stay either slightly negative or slightly positive, but for the most part stayed relatively neutral.

Judging by the graph on the right, most people say they are at the satisfaction level (3), which means "doing good." Many people also chose to respond with a 1 or a 2, which means they are "poorly" or "fairly" satisfied, respectively. Not very many people seemed to report that they were "doing very well" (4) or "excellent" (5), which is most likely because not many people find themselves being happy all the time.

These are our "flirting" pie charts. They depict which segments of the population have flirted with someone over the internet.

We can see that only 17% of the total population in our data have flirted online. We can further break the population down into small groups and see who makes that 17%. We see that 44% of people looking to date and 43% of people who date traditionally have flirted at least once, which is not a big difference. However, there are 62% of online daters who have flirted before. This might mean that dating sites attract people either who are naturally more aggressive in making a relationship move or who are desperate enough that they have tried making a move whenever they have a chance. This flirting behavior might also suggest that online daters are more interested in looking for non-serious flirtationships or hook-ups.

We also used four pie charts to break down how people report their satisfaction based on how they date.

With these pie charts we can see that 75% of online daters are quite satisfied with their lives; whereas when you group everyone together, 54% are quite satisfied with life. Although, when we look at the traditional dates, they also seem relatively satisfied with their lives, so the fact that there are fewer online daters than traditional daters could be making the data look a bit skewed in the favor of online daters, which we found was not necessarily the case.