Romantic Chemistry / How It Was Decoded

At eChemistry.com we didn't assume anything about romantic chemistry. We set out like Galileo, doing experiments to see what there was to learn.

We began by reviewing every piece of research literature we could find on personality and love and and then sent one person on 500 first dates to experimentally determine which factors can be correlated with initial attraction. We spent two years locating, interviewing and psychologically profiling high chemistry long-term married couples to determine which personality patterns make chemistry last. Using interviews and test results from 65,000 users, we then created the world's most advanced personality test, and verified its accuracy by sending a researcher to personally meet and interview hundreds of test subjects.

Using advanced data mining technology we correlated the way real people answer our questions to their real-world personalities. Now this website combines that personality technology with the romantic chemistry research to match you with people with whom you should feel chemistry -- the kind of chemistry which stands up to the test of time.

"Romantic Chemistry" Detailed Definition

In psychology, the word "chemistry" is used when referencing natural mutual rapport between two or more people. When two people have "chemistry" they effortlessly feel a natural connection with each other, even if they've just met. The term "romantic chemistry" refers to a natural mutual romantic attraction between two people which results from a mixture of physical attraction and natural mutual personality-based rapport.

A relationship must meet three criteria to qualify as having true "romantic chemistry":

* Mutual - One sided attraction is simply "attraction", "infatuation" or maybe even "lust." Once any of those terms become what both sides are feeling for each other then it is mutual and satisfies one requirement of "romantic chemistry."

* Effortless - "Learning to love someone" isn't generally referred to as "romantic chemistry." The whole idea is that there's an effortless connection, a natural rapport between people which earns the special moniker. Two people who feel like old friends, even though they've just met, qualify as feeling this natural, effortless connection.

* Sexual - "Romantic chemistry" without sexual attraction is merely "chemistry" or friendship. When two people feel "romantic chemistry" they not only feel naturally like old friends, but they also feel a sexual attraction to each other. The term "sexual chemistry" is used when referring only to the sexual component of romantic chemistry.

We also required a fourth variable in our test subjects: Longevity. Unlike infatuation or even mutual lust, true romantic chemistry doesn't disappear over time. The couples we interviewed giggled like high school crushes whenever they discussed their affections for each other, frequently stopping to say, "Are we being too mushy? It's just that we get so excited talking about this!" even when they'd been married 40 or more years. So while romantic chemistry often exists from the first day, for research purposes we only included couples in the study whose attraction had stood the test of time.

Study #1: Initial Attraction (aka "500 First Dates")

The biggest obstacle to studying the initial stages of romantic chemistry is the substantial number of personal variables involved in the interaction between two people. The challenge is to create a scenario where enough characteristics are held constant for truly scientific experimentation. In other words, chemistry always happens between two people and the best scientific research only has one changing variable. In order to significantly reduce the variables involved, we sent one person on two first dates a week for five years -- which summed to more than 500 first dates -- to determine which traits are involved in initial attraction. In 500 first dates there were multiple instances of attraction and multiple instances of repulsion, all with one individual held constant and only the women (in this case) changing from date to date.

The results from the 500 First Dates study pointed directly toward a crystal clear, specific pattern of factors, which always related in the same way to predict when romantic chemistry is likely to occur. Five hundred samples was more than enough to include a sizable number of personality, socio-economic, physical, birth order and family background variables and the results were that certain factors were dramatically outside their margin of error for predicting initial attraction. Initial attraction is not the same as long-term romantic chemistry, though (or at least we weren't prepared to assume any connections), so a study of long-term romantic chemistry followed.

Study #2: Long-Term Romantic Chemistry

The design of a long-term romantic chemistry experiment is quite straightforward -- it's finding the couples to study that is more difficult. The initial hypothesis (the statement being investigated) is that there is a pattern to personality, background and/or physical traits involved in romantic chemistry which is constant even when the people change. Therefore, though the individual experiences of the pairs are different (since they grew up in non-identical settings with different base personality types), the patterns of who feels chemistry with whom should always be the same. Only very high-chemistry, long-term married couples were used in the first phase of the long-term chemistry experimentation to keep chemistry level constant for data analysis purposes (divorced and low-chemistry couples were added later to expand on the results).

For this two-year study we analyzed many, many high-chemistry, long-term married couples. The results are absolutely conclusive: there is a definite measurable pattern to the traits involved in lasting romantic chemistry and that trait pattern is always the same. That romantic chemistry, like gravity, is a basic law of nature is a surprising result, even for us. And as in Galileo's explorations, the course of every experiment in a natural law starts looking the same once the pattern becomes clear.

The Romantic Chemistry Equation

By approaching this phenomenon from two angles -- from both the beginning of a relationship and after dozens of years of chemistry together -- we were able to identify the full spectrum of the human romantic chemistry experience. A tremendous overlap exists between the results of the two studies: we now understand why chemistry is more prevalent initially between certain people, and why it lasts in specific cases. We also discovered that a study on high-chemistry married couples alone would not have produced complete results, since some traits tend to became more accordant between two people living together over time.

This is the most comprehensive study of romantic chemistry ever undertaken. In typical psychological experiments an 80% correlation rate is considered a success since people are so variant. Our patterns have a nearly 100% correlation over an extensive range of people from incredibly varied backgrounds -- a value which is unusually excellent, and practically unheard of. Though our research and experimentation is ongoing and our technology is always improving, the results of these two studies alone are so convincing and consistent that we have created a way for the world to access this breakthrough now.

Solving the Personality Profiling Problem

Our findings can't change the world until they can be applied to humanity en masse. Although the above research was done in person, personal interview clearly isn't the most efficient method for applying our findings to the world. To this end, we realized the need to create a widely available, fun and easy-to-use test that would accurately measure the traits we have discovered to be integral to romantic matching.

There are several challenges in identifying people's core identities using a web interface, and forefront among them is finding a way to efficiently and accurately profile personalities. Unfortunately, due to the inherently subjective nature of personality and differences in emphasis in mood, attitude, and opinion, people have been found to be inaccurate and inconsistent judges of their own personalities. Many of the techniques that we have developed, and that drive our successes, depend on our ability to extract honest information from our users. What follows is a brief summary of two of the issues faced by creating an online matching system intuitive and flexible enough to be consistently accurate, and how we resolve them to make our personality profile work.

Self Perception
Some people intentionally lie, or choose answers based on what what they prefer to believe about themselves even when they're not true. Some try to manipulate personality tests to give the results they think are "best". Others simply don't know themselves as well as they might claim to.

Emotional Variation and People Changing Over Time
People are molded by their experiences and grow and change as they get older. Our users looking for romantic matches range in age from 18 through 100, so maturation and life experience must be accounted for in final personality evaluations.

The current popular tests do nothing to resolve these issues because they are based on the assumption that these issues don't exist. At eChemistry, we take self-perception, emotional variation and maturation into account. Every aspect of our personality test - the questions used, the way they are phrased, your options for response, the sliding scale interface, and more - is intentional and was designed to account for differences among users.

Personality Testing: Prehistoric Style

Other current personality tests use outdated technology. In the past, test developers relied too heavily on mathematical methods of evaluation and kept personality theory development separate from the evaluation method itself. The result was that the burden of understanding personality theory was placed on the test subject. The clearest examples are the two most common personality tests today: the Myers-Briggs Type Indicator and the Five Factor Model of Personality.

The questions used in the Myers-Briggs Type Indicator were chosen using Midpoint Discrimination Analysis. That means that among the hundreds of questions asked, the ones that were chosen for the final test depended on how far they tended to move the tester's results away from the mean. The result of using Midpoint Discrimination Analysis is that the MBTI measures the consistency of a tester's self perception but is unable to determine whether that self perception is accurate or not.

The Five Factor Model of Personality uses Factor Analysis to distinguish five trait clusters from responses. Once again, this method determines the strength of a person's self-perception, but does not measure the accuracy of the relationship between that self perception and the reality of how that person thinks and acts.

To their credit, both of these tests involved real people in some stage of their development, which was a major step above the previous generation of tests. To their detriment, neither test is able to give any information on whether the tester's self perception is accurate. They operate under the assumption that people know themselves.

This assumption is akin to asking people with smoke coming out of the back of their car to guess at the cause. Some people would guess the transmission, some the engine; some would guess the catalytic converter. If you analyzed their answers to determine which groups of car systems people were guessing parts in, you would have done a factor analysis and have the five factor model solution to car repair. If you asked people three times and found which answers people gave most consistently then you would have the Myers Briggs solution to car repair. Neither of these systems would actually tell you what was wrong with the car though. That would involve someone taking apart the car, experimenting methodically and isolating the issue. That's where eChemistry comes in.

A Revolution in Personality Testing: Sleeping With The Test Subjects

Romantic chemistry is a particular science in which rough estimates of personalities simply aren't good enough. We needed to know who people actually are, including the people with inaccurate self-perceptions, taking into account variations in mood while taking the test, and considering the differences in people who change over time. To continue the analogy from the last paragraph: we don't want people guessing at what's wrong with their car; we want to actually know what the issue is. We needed to move the burden of understanding personality theory from the test subject to the experts.

We had already sent one person on 500 first dates to get real life data on where attraction starts. The obvious next step in developing a serious personality test was to get to know hundreds of people well enough to know for sure what their personality results should be. In March 2006 we sent a profiler out to live with people who had taken the current version of the personality test. That live-in profiling is still going right now. Hundreds of people later, our personality test recognizes the reality of who people are based on their self perceptions.

Where other tests would assume that whatever you answer is an accurate representation of who you are, our test correlates expert live-in evaluation of who you are with the test answers of other people with those traits. So you no longer need to have a forced-choice opinion on whether you're an extrovert or an introvert if you have little or no background in personality theory. The onus for accuracy has shifted from the user to the test creators, whose job it has become to formulate questions that are non-threatening and unbiased enough that people will answer honestly, then to correlate simple reality into complex theory. This necessary information bridge, allowing each end to specialize in that which they actually are skilled in, is possible because of the myriad data mining systems and modern computing power available today. We've combined this capability with modern applications and are providing it now for the public to use.

The Bottom Line

There is not a more thorough study of romantic chemistry and personality testing available anywhere. The fact that we are fundamentally research driven and intent on making our findings available to as many people as possible, at as low a cost as possible, is revolutionary. For single people who wish to find love in a proven, efficient, and exciting way, eChemistry is the next step in the evolution of dating.