7. Subjective data can only be used for limited purposes, such as supporting subjective conclusions or as a tie-breaker when objective data is in short supply or evenly divided.

In the last section we covered why objective data must be verifiable; and when it is not, the data should be classified as subjective. Although of little use for objective conclusions, it can be used in generating subjective conclusions. For example, "lavender is a pretty color" is not data, but restating it as "Sue thinks lavender is a pretty color" makes it into subjective data that may be useful. When at the flower shop, you may use this piece of data to form an argument such as "Sue will like lavender flowers, because she thinks lavender is a pretty color." Since this deals entirely in the realm of subjective, the entire argument makes sense and will probably produce a good decision (purchasing flowers that she will like).

This is the most legitmate use for subjective data, but occasionally it makes sense to bring it in on objective conclusion making. Two such situations are when a decision must be made but you have no objective data, or when the objective data is at a tie. For example, let's say there are exactly two studies on the subject of the healthiness of natural cane sugar versus artificial sweetener when used in foods. The studies are both from credible sources and are published in peer-reviewed journal. Not deciding isn't really a possibility, since most sweet foods and beverages are made with one or the other, and of course you have to eat. So you might call in a subjective piece of data like "I like the idea of things that are natural rather than synthesized in a lab" and use that to break the tie.

Be careful with this! It's easy to start blurring the lines between subjective and objective data - most people do it as a matter of habit. More importantly, it should only be used when withholding judgment and not deciding at all is worse than any of the options. (More about this on Tenet 9.)