One of the main goals of science is find the relationship between variables. Variables refer to anything that can vary. Since everything can vary, everthing can be a variable.
To describe anything, we have to measure it in some way. To say how tall Mary is, we have to measure her height. Since height will vary from person to person, height is a variable.
There are two important characteristics of measurement: Reliability and validity. Reliability means you get the same result every time you measure. For example, if you measure Mary on Monday, you will get the same result you got on Sunday. A measure that changes each time you measure is an unreliable measure.
Validity means a measurement is measuring what it is supposed to measure. If you measure someone’s lifespan by how long a line on their palm is, that is an invalid measure because lines on your palm have nothing to do with how long you live. However, such a measure is reliable because you get the same thing every time..
People often conflate (equate) validity with reliability – they are not the same! However, we can say this: A measure which is not reliable cannot be valid. Sometimes this fact can be used to determine whether something is true.
For example, I was walking through the mall one day and saw a booth where you could ostensibly (look it up) have a computer analyze your personality for $2.00. I thought, Why not? I gave the two ladies my money and they asked me to write my name on a piece of paper which they then fed into the computer. There was a wild blinking of lights on a panel and shortly after my analysis spewed out.
I was gratified to learn that I was “kind, but sometimes a little selfish,” “fiercely loyal to my friends,” and “sometimes held back by my fear of taking risks.”
Well, that was actually a quite accurate measurement of my personality. Or was it? I decided to do a reliability check. After all, it was only $2.00 (this was before the stock market crashed).
The second printout read that I “often took unnecessary chances,” “was generous, but sometimes too generous,” “was intelligent but sometimes made stupid mistakes,” etc., etc.
What happened? How could the two analyses be different? It was the same person both times. It was a classic case of unreliability.
I confronted the ladies about this and they said that I may have written my name slightly different the second time. “And what?” I asked. “Changed my whole personality in two minutes by writing my name a little different?”
As a missionary for critical thinking assigned to the world, I felt obligated to stick around for a while and explain to those interested patrons stopping by that the whole setup was a scam because the computer didn’t give reliable results. Judging from the looks the women in the booth gave me, I don’t think they appreciated my zeal for truth.
The point is that a measure that is unreliable cannot be valid.
Using this principle, you can also demonstrate the invalidity of astrology. Sometime check your horoscope in two papers. If they’re different they’re not reliable, if they’re not reliable, they can’t be valid.