I've posted in the past about using Enneagram tests in general. This post is about the methodology used in a test and it's purpose. Let's take a look at the methodologies used by the different authors and teachers in their tests (and then I'll get to my methodology and reasoning).
There's the forced-choice test. You're given two statements to choose from with each representing a different type for each question. So, you're basically forced to pick one type or the other. Whether or not you feel strongly or ambivalent about either one, you must choose. The total score is added and the type you chose most often is the winner (followed by next closest scores which can be considered possible candidates if the winner doesn't fit). This is the method used with the RHETI or Riso-Hudson test. There are sample versions of the RHETI on various websites (just google "sample RHETI" to get a list) but the full version at Enneagram Institute will cost you.
There's the Likert style test where you have responses to a statement along the lines of strongly agree, somewhat agree, neutral, disagree, strongly disagree. Each response adds a different point total to the type being described in the statement. Add up the scores and again you have the winner followed by close candidates. Most tests out there that are anonymous or from people you may have never heard of use this method. Probably because it's the easiest test to make given the test software available on the web.
Then there are keyword tests where you choose words that you identify with and under the covers they add points to the type they represent. Again, total the points and find the winner and alternate candidates. Jerry Wagner seems to use some variation of this method on his WEPSS test. Although you have to pay for that test, he offers a brief alternate version online called Ennea-Graph.
The Fauvres at Enneagram Explorations use a unique approach choosing image collages that you relate to called enneacards. While you have to pay for the full tests, they do offer a free sampler.
David Daniels created a test based on paragraph selection which can be purchased online or found in his book The Essential Enneagram. Each paragraph describes key aspects of a given type and you simply contrast and choose between those different types. Now you might wonder how that's different from just reading descriptions and choosing your type yourself. Well, it's sort of a divide and conquer strategy. You narrow the choices down in stages and then compare fewer types as you go.
How do I see it? Ideally, you can just go through the type descriptions in a book or elsewhere and recognize which one best fits for you. The problem is that the descriptions are based on author interpretations and can vary greatly. Not only that, but many of the descriptions attempt to describe personality traits and characteristics that don't match 100% for an individual (i.e., you might not identify with all the characteristics of any single type and you might identify with some characteristics of one type and some of another type). Using a single word or brief statement can fall into the same trap as well if they're describing personality characteristics.
The way I see it is that people either don't know anything about the Enneagram types and just want to take a test to quickly determine their type out of passing curiosity or they're confused by all the different interpretations of the types they've read. For the people who just want a passing idea of what type they might be, no test will suffice because they are simply not that accurate. It's up to the individual to do some research into the type in order to determine for themselves what type they are. Those confused by all the different descriptions need something that helps cut through all the noise and offers the core aspects of each type to make it easier to compare and contrast the differences between each type.
The approach I'm using in my tests is to reduce the choices down by clearly contrasting the types. This is more of an adaptive style test where the goal is not to add up the points and see who the winner is but to narrow down choices based on previous choices until a likely candidate appears.
I've created two tests so far that do this (and both are free right now). One test narrows down type and wing. The other test narrows down Tritype and the associated wings. Maybe when I finish the book I'm working on or as part of an online discover your type training, I'll create an instinctual stacking test as well.