Where do personal observation, experience, and values fit in scientific research?

What is the appropriate role of personal observation and experience in scientific research?
I wonder about this a lot because many of my research ideas, as well as my posts, come from things I have observed or experienced.  I have mixed feelings about this.  
On the one hand, life has been a rich source of ideas, and with so many bubbling up, some will turn out to be original and useful.  Personal experience also keeps me passionate about research.  I love mastering new areas of knowledge, but without a personal connection, how could I choose a single one out of the nigh-infinite numbers of fascinating systems out there to research?  Sheer fascination, and a desire to make neurodiverse people's lives better, keep me going on days when all my research tasks are deadly dull.  
On the other hand, objectivity defines science.  The law of gravity affects everyone and everything according to specific mathematical laws, no matter how we feel about it.  In a person who can see colors, the neurons fire in the same pattern when they see blue no matter how blue looks to them, and while everyone may not experience the same blue, their brains fire the same way when looking at it.  Yet, while I clearly observed how sensory processing problems interfered with my brother's functioning, I could not necessarily conclude the same about every child with Asperger's, because it might not work the same way for someone else.  There is too much individual variability, too much random chance, and most of all, my brother might fall into a category so rare that a study will never include enough people like him to reach statistical significance.  My honors thesis probably failed for this reason.  I personally knew several children and adults who appeared to have intuitive logical ability, and I wanted to establish this ability in a population of gifted children and learn something about its basis.  However, the population of 5-8 year old children with intuitive logical ability may be so small that an average-sized study like mine could not sample enough of them to find out anything significant about how they think.
Most of all, I worry that relying on personal experience for ideas will interfere with my objectivity.  If I "know" that sensory processing has cascading effects on social functioning from such a source, will I ignore good evidence against this hypothesis, and let pass flawed studies that support it?  I hope simple awareness of this possibility will keep me honest.
But there is danger in confusing emotional detachment with objectivity.  When I read the cold, harsh prose in autism studies coming from the Yale Child Study Center, or Simon Baron-Cohen and his colleagues, I wonder, "have these people ever met a person with autism?"  No one who loves a person on the spectrum, I think, could possibly write about them in such a way.  Yet instead of objectivity, they have their own set of anti-autistic biases that may interfere with doing good science (see also this and this).  Some would say Michelle Dawson could not possibly conduct objective research because she is autistic herself, and Morton Ann Gernsbacher could not because she has an autistic child.  Yet I find none of the same reasoning and methodological flaws in their work that cripple so many other autism studies.  So, maybe having a brother on the spectrum, and a deep commitment to conducting ethical research, won't automatically disqualify me from doing good science.
I've just swung back and forth repeatedly, but haven't actually landed anywhere, so it's time for a fresh perspective.  What do you think--is it possible to do valid, objective science if you have personal investment and experience in the matter?  If so, how can you use your experience and emotions to make you more objective, rather than less?  Whether you're a scientist or you just like reading science, I'd love to hear from you.