Tests as measuring instruments

[WasGSOHM]

If you take a test and you get zero, has this test measured anything?

Same question for when you get 100%

I picked this up from a course my ex-wife took.

This thread title contains a hint.

 

[zbang]

For zero- sure, assuming the instrument is operating correctly. For 100%, it depends on how the instrument behaves when over-range.

 

[Buck Parrish]

The answer can be in relations to a lot of different scenarios. But I used electrical for obvious reasons
1) No - If you put your leads on and get 0 . It may be hot but no neutral.
2) Yes - If you get a 100 percent . You're done.

 

[WasGSOHM]

The range of IQs is 90 to 220 something.
You have some inkling what voltage you'll be seeing.

If a guy gets 100% on his test, make his next test twice as hard.

If a guy gets 0%, make his next test half as hard.

I'd say if your meter reads .01 you know it's .01, or it's less. Switch ranges or switch meters.

 

[infinity]

Does this have anything to do with electricity? The way I read it, no.

 

[Hv&Lv]

Zero error question?
as in zero on a thermometer is an arbitrary point, or even an IQ test, 90 or 100 is just an arbitrary point on a scale.
A true zero, 100 test would be a “go-no go” test.

zero volts on a voltmeter means get another meter...

 

[WasGSOHM]

Does this have anything to do with electricity? The way I read it, no.

It has to do with test making, so it has to do with the forum category of Education.
Knowing how tests are made may help with your test-taking strategy.

It also has to do interpreting
meter readings/test results.
Tiny voltage readings may have meaning or may just mean noise. I'd check them out.

If this post helps a member or some viewer somewhere, good.

If not, please handle IAW forum rules.

I need to explain relevance better than I've been doing.

 

[mikeames]

Its philosophical. The answer is YES. A zero tells you something. It tells you that the person does not know what is being tested. Test results of any kind are relative. You know the content of the test. So a zero speaks to what is tested as written/ asked. To put it in infinite context of all knowledge is silly or philosophical. Much like measuring voltage. Everything will have some voltage relative to some other thing but that does not have much value. You only care about the voltage relative to your system/ TEST. You don't change the test to match the voltage. Range on a meter is still the same test with different resolution. A 100% means the person knows the content you tested on as written.

Everything is relative, lets not go down that hole.

 

[kwired]

With the meter how to read the results is dependent on what you are looking for in those results.

If testing for volts and you get zero, you may want to test something else that you expect to get some other reading on just to confirm the meter hasn't malfunctioned. If the test is just for confirmation of energized/non energized, any reading within nominal range of what you expecting to be normal probably is a valid "energized diagnosis", but then again there is high impedance meters that could still fool you if you don't understand the situation well enough.

 

[LarryFine]

This reminded me of a book about mechanics and physics for kids I read as a kid.

It showed a clown trying unsuccessfully to lift a chair that was nailed to the floor, with a feather in the seat. Then a second clown effortlessly picked up just the feather.

The book then asked who did more work, the answer being the second clown, because the work actually done was dependent on the results obtained by the effort applied, not on the effort alone.

Your question suggests that the success of a test is dependent on the results (presuming the results are accurate). Why wouldn't a zero reading on a de-energized circuit be considered a successful test?

 

[kwired]

This reminded me of a book about mechanics and physics for kids I read as a kid.

It showed a clown trying unsuccessfully to lift a chair that was nailed to the floor, with a feather in the seat. Then a second clown effortlessly picked up just the feather.

The book then asked who did more work, the answer being the second clown, because the work actually done was dependent on the results obtained by the effort applied, not on the effort alone.

Your question suggests that the success of a test is dependent on the results (presuming the results are accurate). Why wouldn't a zero reading on a de-energized circuit be considered a successful test?

I'd say most cases it is a successful step in the overall test. If you want to be certain there is no voltage, you at least need to verify the meter will successfully show a voltage reading where there is indeed a voltage.

 

[LarryFine]

I'd say most cases it is a successful step in the overall test. If you want to be certain there is no voltage, you at least need to verify the meter will successfully show a voltage reading where there is indeed a voltage.

That's why the best method is to first test the tester on a known source, take your test reading, then test the tester again.

That way, you know the tester didn't just happen to fail after the first tester check and give you a false no-power reading.

 

[paulengr]

Saw this in a TV show recently where a student took a test and got exactly 0%. The teacher said it’s mathematically impossible. On a multiple choice test if you just guess you should get about 25% and the fact that she got every single question wrong proves she knew the right answers every time.

So 0% means you cheated.

But the concept gets into whether a test measures anything if it’s too easy or too hard. Kind of like when an ohm meter reads 0 because the actual resistance is too small to measure or open because it’s too big. Last week a plant needed to know if their soft start was any good. The guy had a Fluke 87 V. It read open but the correct reading was about 2 megaohms. On another job the electrician attempted to read the resistance of a 200 HP motor with a multimeter and say it was 0.2 ohms which might be his lead resistance. The correct reading was 35 milliohms...not something his meter could read.

 

[mikeames]

The teacher said it’s mathematically impossible. On a multiple choice test if you just guess you should get about 25% and the fact that she got every single question wrong proves she knew the right answers every time.

That's depends on how many questions, and how they were written. A 5 question multiple choice test is entirely possible to fail with a 0%. Add in human bias on the teacher and the test taker and its possible to get a 0% even with 20 questions. Especially if a student thinks they know a concept and has it wrong. They may be working on every problem exactly wrong. Randomness goes out the window there. A poorly written test that incorporates the same concept in multiple questions is one example.

 

[Hv&Lv]

The teacher said it’s mathematically impossible. On a multiple choice test if you just guess you should get about 25%

I had a friend that took the FE that would dispute that...

 

[Flicker Index]

Saw this in a TV show recently where a student took a test and got exactly 0%. The teacher said it’s mathematically impossible. On a multiple choice test if you just guess you should get about 25% and the fact that she got every single question wrong proves she knew the right answers every time.

No, that's BS. You can absolutely setup multiple choice test so that each question has one answer that fails sanity test. You can certainly have answers where you KNOW for sure which one is wrong and you purposely go for the wrong one and you can specifically go around selecting answers that fail the reasonableness test.
For example, I give you a ball that fits in your pocket and give you the exact diameter.

What is the internal volume? You have to know how to calculate the volume of a sphere to get the right answer...
1.) 15.9oz
2.) 22oz
3.) 18 oz
4.) 2 gallons

but you know #4 fails the sanity test, because the volume inside even the largest pocket would not hold two gallons.

 

[LarryFine]

but you know #4 fails the sanity test, because the volume inside even the largest pocket would not hold two gallons.

Is that two gallons of nails or two gallons of feathers?

 

[Carultch]

Saw this in a TV show recently where a student took a test and got exactly 0%. The teacher said it’s mathematically impossible. On a multiple choice test if you just guess you should get about 25% and the fact that she got every single question wrong proves she knew the right answers every time.

The teacher is confusing possibility with probability. 0% on a multiple choice test is possible, if you are extremely unlucky. Suppose it had 10 questions with 4 answer choices each. The probability that you get zero, from random guessing is 0.75^10 = 5.6%. That's about the probability of rolling the automatic loss in Craps on a standard pair of dice, i.e. a roll that is either "snake eyes" (2) or "boxcars" (12). Set up 100 students to do random guessing on this test, and it is realistic to expect that five or more of them will get a zero.

It also could be possible, if there is a red herring written in every question, that increases the odds of picking a wrong answer. So much that it is at least equally likely for a student knowledgeable about the subject material to get the wrong answer, than to get the correct answer. As an example, "There is a jar with 5 marbles, and you take 3 away, how many marbles do you have?". One might think a run-of-the-mill subtraction problem and the answer is 2 marbles. But it is a trick question, because "you" are the one who took them, so the correct answer is 3 marbles.

 

[LarryFine]

"You're driving a bus . . . "

 

[zbang]

The teacher is confusing possibility with probability. 0% on a multiple choice test is possible, if you are extremely unlucky. Suppose it had 10 questions with 4 answer choices each. The probability that you get zero, from random guessing is 0.75^10 = 5.6%.

If you actually know the answers, getting 0% is a cinch; just choose the wrong ones .