Open Book Tests?

[Pam B]

What is the point of an open book test?

 

The only thing I can come up with is good referencing skills! LOL :D But, I'm pretty sure there must be more????

[Teachin'Mine]

I've always wondered the same thing! :) But I never complained when we had one. :D

[Dana]

One reason would be that then you can ask a question that goes deeper in depth for analysis instead of just basic knowledge.

 

For instance, with English, you could allow the book and then you'd expect supporting quotes.

 

Again, it's only going to be an effective test if it's testing higher-level skills and not just regurgitation of facts (although in a larger class, that can be done just to see if you learned the facts and can find them quickly).

[Gwen in VA]

In my science and engineering classes, open book tests helped the tests to be about content, context, and analysis, not just memorization.

 

In real life, engineers have access to formulas and often use reference materials to check that they are using the correct formula. The real issue is when to use which formula, and what assumptions you can or need to make to use which formula. This process can be a lot more complicated than it sounds, and there is an art to it.

[regentrude]

There are subjects that are not about memorizing - such as physics. With the open book, you may have access to formulas and to problems of a similar kind, but the actual test problem will be different form the examples. So it can be tested whether the students understand the underlying concepts and are able to generalize what they have learned to an unknown system. Which is far more valuable than having them spend their time memorizing the formulas.

I can imagine an English open book test as well: you might have access to literary terms, or characteristics of s certain genre, but then be confronted with an unknown text which you have to analyze.

[Pam B]

Thank-you!

This really does shed some light on the process!

[memphispeg]

I am in my 1st year of homeschooling (jr. year - high school) and I have been doing open book tests for Math and Bio and maybe English to build my dd's confidence. Most of the tests I have been giving have required analysis and synthesis of the material. I looked at the OCW/MIT science and math tests and this is what they require (the tests looked really fun and interesting). We will eventually close the books but I'd like to get her thinking and writing skills going before I do that.

[yinne]

In college I had a business statistics course that I was allowed to use the book for, but only because I was a business major. If you were a math major taking the same class (but called a math statistics course) from the same professor you were not allowed to use the book. We always figured he thought business majors were not as smart as math majors -- but we weren't going to disagree with him.

 

Of course this was the same professor that handed out graded tests in order of grades. You knew you were toast if your paper was returned last -- especially as he made comments as he returned them!!

 

Yvonne in NE

[oldskool]

In my science and engineering classes, open book tests helped the tests to be about content, context, and analysis, not just memorization.

 

In real life, engineers have access to formulas and often use reference materials to check that they are using the correct formula. The real issue is when to use which formula, and what assumptions you can or need to make to use which formula. This process can be a lot more complicated than it sounds, and there is an art to it.

 

 

This is the explanation that I have been given as well. That in real life people will have access to material and it is knowing how to find it and use it in a timely manner that is most important.

[TravelingChris]

In my Government class, they had closed book tests. In Economics, it was open book test. Guess which class had more people getting A;s? It was government. Why did I do an open book for economics? Because I didn't want the test to be about who memorized the GDP formula or some other formula. I wanted it to test understanding. While all kids did improve in understanding, only about a third really mastered the learning.

[EKS]

The hardest tests I ever took in college were for the second quarter of a three quarter long biochemistry sequence about enzyme kinetics. They were open note, but you really had to know what was going on to do well and if you had to look up every last thing, you were doomed.

[StephanieZ]

Open NOTE tests teach good note taking skills. . .

[Ester Maria]

There are subjects that are not about memorizing - such as physics. With the open book, you may have access to formulas and to problems of a similar kind, but the actual test problem will be different form the examples. So it can be tested whether the students understand the underlying concepts and are able to generalize what they have learned to an unknown system. Which is far more valuable than having them spend their time memorizing the formulas.

On the high school level, I wouldn't use open book tests for sciences or mathematics. IMO, it's a completely different thing if I allow them to use Bronstein, or similar handbooks for sciences, which contain all the formulae the might need - and if I allow them to use the actual texts they've studied from with solved examples of problems. That's a bit like giving them half of the test solved, only with different numbers - what's the point? It's like a guarantee of a passing grade to anyone intelligent enough to realize it's the same thing with changed numbers, and to directly copy the answers for the easiest theoretical questions.

 

With my daughters, I have the following system: they don't have to know the formula, but they need to understand the process. I don't care if the formula looks the way it "should" look, or if it's something like Ax^2B=C, as long as they explicitly write what's A, what's B and what's C, and if the relations between the phenomena that the formula describes are understood properly. I did that a bunch of times on my own science tests in high school when I couldn't recall the exact formula or I couldn't recall the symbol for something, but I knew in what relations the phenomena studied were. For lenghtier formulae, I'm even willing to allow those to be openly used - but NOT from a book. They can be written on a special piece of paper, or consulted from a handbook which contains only various formulae (no solved examples or elaborate explanations). For chemistry, periodic table may always be used of course, as detailed as you wish.

 

On an open book test, a relatively intelligent student that knows only 50% of the material can easily get up to 80% of the grade - all it takes is to vaguely know what it's about and to know how to use a book. Also, a student that knows maybe 40%, which wouldn't be enough to pass, may reach the limit of 60% only by the virtue of solving same problems with different numbers or answering to the easiest theoretical problems. It'll be a passing grade for somebody who normally wouldn't pass. And while I agree that it's almost impossible to get an A without actually knowing the content, a grade can be certainly "artifically lifted" this way. I also find the high school content, by its dififculty, simply to not justify the need for open book tests. Those aren't classes for engineering majors, it's quite elementary Physics/Chem/etc.

 

If you know something, you know it, open book or not. But if you don't know it properly, but only have a vague idea of what's going on, an open book can artificially raise your actual grade by even more than one grade - which is what you're trying to avoid. Access to formulae is one thing, but the actual book or notes... No way. Not at this level of complexity.

[regentrude]

On the high school level, I wouldn't use open book tests for sciences or mathematics. IMO, it's a completely different thing if I allow them to use Bronstein, or similar handbooks for sciences, which contain all the formulae the might need - and if I allow them to use the actual texts they've studied from with solved examples of problems. That's a bit like giving them half of the test solved, only with different numbers - what's the point? It's like a guarantee of a passing grade to anyone intelligent enough to realize it's the same thing with changed numbers, and to directly copy the answers for the easiest theoretical questions.

 

While I do not do any open book tests at all, what you describe would be a

poorly designed open book test.

If you use the same problem with different numbers, it's a no-brainer.

However, I can design a physics problem that has similar concepts, but looks completely different from the worked examples, so that true understanding is tested.

It is hard work to design a good open book test. Just as it is hard work to design a good multiple choice exam that really tests understanding - much harder than composing an open question test.

[Ester Maria]

I can design a physics problem that has similar concepts, but looks completely different from the worked examples, so that true understanding is tested.

Yes, but why would the book be needed for that? Why not just give a list of formulae, but zero explanations and zero solved problems? Even with a well-designed test, I find that this is a lot more elegant solution than just allowing the book.

It is hard work to design a good open book test. Just as it is hard work to design a good multiple choice exam that really tests understanding - much harder than composing an open question test.

I agree - which is why I don't do open books OR multiple choice exams OR "fill in the blank" exams. Because they very rarely test knowledge, they usually test adaptability (which is not bad, each test needs a bit of that; but when concrete knowledge is not tested in addition to it, I don't think it's good) and familiarity, memory, the ability to recall or quickly rule out some options, etc.

 

I usually use essay-type exams for theory, and a set of problems with MAYBE open-formulae, but like I said, I prefer them to understand how things are related and "invent" the formula if needed right there on the test.

[regentrude]

I agree - which is why I don't do open books OR multiple choice exams OR "fill in the blank" exams. Because they very rarely test knowledge, they usually test adaptability

 

Again, a well designed multiple choice exam does test real knowledge.

All it takes, for instance, is to ask a question that must be answered by solving a fully worked problem, and just offering them four answer choices without any easy way of eliminating wrong answers. Students usually do more poorly on those, because they have to do the same amount of work as on a test with open ended questions, but will not receive any partial credit for steps in their solution if it is not entirely correct. (And most students in my classes live on partial credit)

If I make the wrong answer choices ones that would be obtained with the most commonly made mistakes, I can write a very hard multiple choice exam (without resorting to any trick or mean questions), that tests true knowledge and understanding.

I found multiple choice exams a very valuable solution at the end of the semester to eliminate any haggling for points ;-) - while all my normal exams are full problems with partial credit, my Final is multiple choice. Guess on which exams the students get better grades.

[Pam B]

:bigear: Amazing insight!

[Ester Maria]

Again, a well designed multiple choice exam does test real knowledge.

All it takes, for instance, is to ask a question that must be answered by solving a fully worked problem, and just offering them four answer choices without any easy way of eliminating wrong answers. Students usually do more poorly on those, because they have to do the same amount of work as on a test with open ended questions, but will not receive any partial credit for steps in their solution if it is not entirely correct. (And most students in my classes live on partial credit)

If I make the wrong answer choices ones that would be obtained with the most commonly made mistakes, I can write a very hard multiple choice exam (without resorting to any trick or mean questions), that tests true knowledge and understanding.

I found multiple choice exams a very valuable solution at the end of the semester to eliminate any haggling for points ;-) - while all my normal exams are full problems with partial credit, my Final is multiple choice. Guess on which exams the students get better grades.

I get what you're saying, but there's still a theoretical option of guessing one's way through, which doesn't exist on open ended questions. Unless you give negative points ;) (so students don't dare to guess), the result can still be largely affected by guessing - that's why I oppose such tests, since I'm looking for something which minimizes guessing: if you require full process (and DO NOT give partial credit unless the WHOLE thing is correct), full answers, etc., one cannot guess - it's either you know or you don't know.

[regentrude]

Point taken - I understand what you say about not liking the guessing.

 

Let me just share my experience of nine years of teaching and having several hundred students take a multiple choice final exam:

the ones who know only little bit and then guess on part of the questions of our four option test (30 questions in 2 hours) score around 50% which is an F. OK, it is a higher F than the random 25% (after all, they did manage to work out some of the problems), but there is no such thing as a "good" F and a "bad" F - it's failing :-)

I have never had anybody guess their way to a B.