"The light fixture that wears out quickly is in a small box-like space covered by a thick piece of glass, so it must get hot in there- no wonder the bulb burns out so often. I will try switching the glass."

You may need to do more than that, particularly with a high-wattage bulb. What matters is how hot the bulb gets.

"I have a dimmer switch hooked up to my bedside lamp, which means that the bulb receives lower voltage, and thus lasts longer."

A CFL that isn't meant to be on a dimmer will have a short and dramatic life when used with one.

Running an incandescent on even slightly lower than design voltage will dramatically increase its life - and dramatically reduce its efficiency. For example, if you run an incadescent at 80% of its design voltage, it will last 20 times its normal-voltage lifetime. But the efficiency (amount of light per watt) will be about half of its normal-voltage value.

"Now if only I knew why the lightbuln in my kitchen light seems to burn out so often- it's in a well-ventilated location and not on a dimmer, though we do turn it on and off five or six times a day. Is that considered "frequent"?"

No.

However the *fixture* may not be well ventilated. What matters is how hot the bulb gets.

"Should I use an incandescent? The electrical system of the (rental) house is definitely a bit shady, so I wouldn't be surprised if it's not receiving consistent voltage."

Consider replacing the fixture and/or writing date-of-installation on the bulbs. Relying on memory can be tricky.

Ceiling lights and those near doors are particularly susceptible to failure from vibration, due to door-slamming and "heavy walkers" above. Consider rugged-service bulbs for such applications (they do exist, but watch out for left-hand threads).

"For those that claim no link from vaccines to autism, can you provide me with just one conclusive study?"

A study that conclusively proves there is absolutely no link? No such study exists - because, by definition, it cannot exist.

The problem is that, in most cases, a negative cannot be proven. Only a positive can be proven.

This is why reputable scientists will say "we have no evidence of a causal link between A and B", but will not say "A does not cause B" except in special, specific cases.

So what is needed is a conclusive study that shows a definite link. That means a large sample group, controls, repeatability, and much more. That's very difficult to do with humans because of all the variables.

--

The plain and simple fact is that nobody really knows what causes autism spectrum disorders. It's not even known if they are a single condition that manifests itself in various ways and degrees, or if they are a number of different conditions with similar manifestations. And while there seems to be a genetic link to ASDs, it is not known whether it's purely genetic, a genetic sensitivity to an environmental factor(s), or some combination.

We're used to getting solid cause-and-effect answers to problems. Autism provides the ultimate frustration by (so far) defying that.