"Birds of a feather flock together".
If your family is composed of outliers* they likely know other outliers. In syndicated research, you will be counted but users of the research will ignore you.
The classic example of the outlier is the person who will consume garlic flavored ice cream again after once trying it. But if research shows less than 0.1% of consumers like that flavor, and you only have space in the store freezer cases for 31 different flavors, you will choose not to stock that flavor.
Outliers tend to associate with other outliers. So it is likely that members of your family gravitate towards other people who don't have mass, mainstream preferences.
Outlier: A precise term for those people who are way, way outside the range of behaviour. For example, if we find that 90% of a universe falls within a range of +/- 10 on a scale of 100, then someone who is off by 50 on that scale is an outlier... because, when graphed, that person lies way outside the responses for the 90% who are in that narrow 20 point spread. We have a variety of statistical techniques that we use to identify outliers, and in proprietary research it's usual to look at results with the outliers purged from the dataset as those influences will actually have a negative effect on pleasing that 90% "vast" majority.