Teaching Proofs in Geometry - What I do.

This is the second year that I've had a standard geometry class to teach. The other times when I've taught some of the same topics, it has been in the context of integrated curricula, so there wasn't too much emphasis on proof. When the time came last year to decide how teaching proofs fit into my overall teaching philosophy, it was a new concept. I've seen some pretty amazing teachers have great success in teaching it to students who subsequently are able to score very highly on standardized exams. I'm in the fortunate position of not having to align my proof teaching to the format on an exam. As a result, I've been able to fit what I see as the power of proof-writing to the needs and skills of my students in the bigger context of getting them to think logically and communicate their ideas.

As a result, my general feeling about writing proofs is as follows:

  • Memorizing theorems by their number in the textbook is less important than being able to communicate what they say.

I'll accept 'vertical angles theorem' but fully expect my students to be able to draw me a quick diagram to show me what the theorem really says. This is especially important for international school students who may move away to a new math classroom in another part of the world in which 'Theorem 2-3' has no meaning. I won't ask students to state a theorem word for word on an assessment either, but they must know the hypothesis and conclusion well enough to know when they can apply it to justify a step in their proofs.

  • Being clear about notation and clear connections between steps in proofs is important.

Since the focus of my geometry class is clear communication, correct use of notation is important. If angle A and angle B are congruent, and the measures of these angles are then used in a subsequent step of the proof, it needs to be stated that the measures of angle A and measures of angle B are congruent equal. I'm not going to fail a student for using incorrect notation in a proof if the rest of the logic is sound, but a student will not receive a perfect score either if he/she uses congruent angles interchangeably with their measures.

  • Struggling and getting feedback from others is the key to learning to do this correctly.

I don't want my students memorizing proofs. I want them to understand how logic and theorems applied step by step can prove statements to be true. Human interaction is key to seeing whether a statement is logical or not - I like taking the 'make-it-better' approach with students. If a student says angle A and angle B are congruent, and that statement is not given information, then there needs to be some logical statement to justify it. In all likelihood, there is another person in the classroom that can help provide that missing information , and it won't necessarily be me. As I wrote in a previous post, it was tough letting the students struggle with proofs in the beginning, but they helped each other beautifully to fill in the gaps in their understanding. This makes it hard for students that are used to being able to see a thousand examples and get it, but since that isn't my intent for this course, I'm fine with that.

My progression for teaching proofs starts with giving the students a chance to investigate a concept and predict a theorem using Geogebra or a pencil and paper sketch. I like using Geogebra for this purpose because it instantly lets students check whether a property is true for many different configurations of the geometrical objects.

As an example, the diagram at right is one similar to what my students made during a recent class. The students see that some angles are congruent and that others are supplementary. They can make a conjecture about them always being congruent after moving the points around and seeing that their measures are always equal. This grounds the idea of writing a proof in the idea that they know that if parallel lines are cut by a transversal, then alternate exterior angles are congruent. There's no failing in this if the activity has been designed correctly - students will observe a pattern.

The work of writing the proof doesn't start here - usually some work needs to be done to get a complete conditional statement to be used as a theorem. When students suggest hypotheses for the statement, and it isn't as complete as it needs to be, I (or even better, other students) play devil's advocate and construct diagrams that might serve as counterexamples for the entire statement NOT to be true. Students might suggest 'if two lines are intersected by a third line, then alternate exterior angles are the same'. If I've done my job correctly, students will (and at this point are) catching each other on using congruent rather than the same, and not saying that angles are equal. This is a great spot for the students that love catching mistakes (though often don't catch their own). Until students are comfortable writing the theorems using precise and correct mathematical language based on their observations, writing the proofs themselves is a huge challenge.

I balance the above activities with another introductory step in writing proofs. I'll provide the statements in order for the proof and ask students to provide the reasons. This works well because students seem to see coming up with the statements as the tough part, and the reasons come from a menu of properties and theorems that we've put together previously in class. I don't like doing too much of this as it doesn't require as much social interaction aside from "is this the right reason?" from students as the rest of proof writing.

The final step to writing proofs comes in the form of returning to a diagram like that above. If students are proving the statement "if parallel lines in a plane are cut by a transversal, then alternate exterior angles are congruent", I expect them to draw a diagram (on paper or on Geogebra) showing parallel lines cut by a transversal. I tell them to pick an exterior angle and give it/find its measure. Then they need to go step by step and find the other measures of the angles using only theorems we know, and NOT using the statement we are trying to prove. (We call this the 'cheap' way.) My way of prompting this development is by asking questions. If a student is sitting and staring at angle 3 in the diagram, I can ask about another angle he/she knows is congruent to that angle. A student will invariable state a correct angle just from having a correct diagram, but this is the important part: the student MUST be able to identify in words what theorem/postulate allows the student to say that the other angle is congruent, either verbally or in writing.

The key thing to show students at this point is that there are MANY ways to make this process happen. Some will see vertical angles right away, and say that angle 3 is congruent to angle 2 because of the vertical angles theorem. This then leads to seeing that angle 2 is congruent to angle 1 by the corresponding angle postulate, and then the final step of using transitivity to prove the theorem. Some students will jump from angle 3 to angle 2 (vertical angles theorem), then angle 2 to angle 4 (alternate interior angles theorem), then angle 4 to angle 1 (vertical angles theorem again). Having students share at this point the many ways of doing this is crucial - letting them justify which angles are congruent using concrete values for the angles, and justifying each step with another theorem, definition, or postulate is the important part. Once they have done this, I let them work together to write the full proof using the concrete road map. They don't get it right the first time, but having the real numbers as an example grounds the abstraction of the idea of proof enough for students to see how the proof comes together.

The weaker students in the group need one extra step sometimes. I let them fill in all of the angles in the diagram first using what they know - this part, they tend to be pretty good at, and I don't flinch when they use the calculator to do the arithmetic since some need that to be successful. Then we hop from angle to angle and the student must explain using the correct vocabulary why the angles are congruent or supplementary. In keeping this as an exercise in concrete numbers, I've had some success in these students (and the ESOL students) using the correct vocabulary, even if they are unable to write the proof completely on their own.

I started to see the dividends of this progression this week, and I am really pleased to see how far they have come in being able to justify their statements. The only thing we did need to work on was how to structure an answer to questions that ask students to make a conclusion based on given information and the theorems they know. This was in response to using converses of the parallel line theorems to show that given lines are parallel. To help them with this process, I gave them this frame and set of examples:

I was very impressed with how this improved the responses of all students in the class. We had some great conversations about the content of student conclusions using this format. In diagram (b), two students had different conclusions about why lines CF and HA are parallel, and there was some really great student-led discussion in explaining why they were both correct. I forced myself to listen and let their thinking guide this discussion, and I was really happy with how it worked.

This year's group still is not super thrilled about having to write proofs, but they are not showing the outright hatred that the last group was showing at this point. I have been emphasizing the move from concrete to abstract much more with this group, as well as showing that the proof is really a logical next-step from reasoning how two angles with definite measures relate to each other in a diagram.  If nothing else, students are already better communicators of their math thinking in comparison to the first day of class when there was plenty of wild gesturing and pointing to 'that thing, yeah' on the whiteboard. Continuing to develop this is, I believe the real goal of a geometry class and not the memorization of theorems. My next step is to include the statement writing process as the first step in solving an algebraic problem. Many students are still throwing the dice and either setting algebraic expressions equal to each other, or adding them together and setting them equal to 180 because that's what they did in their other classes. I am trying, trying, trying to get them out of this habit.

I hope that in sharing my process, others might get ideas on how to either make a certain geometry teacher better (me) or to enhance what is already going on in other classrooms. If any readers have suggestions on how to improve this as time goes forward, I am most thankful for any and all advice you can provide.

2 thoughts on “Teaching Proofs in Geometry - What I do.

  1. Hello. I am a substitute teacher who has taken over for one who is out long term and I am trying to teach students proofs in geometry, among other topics. I certainly like your above recommendations and wonder if you might have knowledge of a source book related to teaching proofs that can supplement my students. Our book is not particularly helpful and the biggest problem our students have is the universal one... Where does one begin right after listing the givens. I suggest drawing a diagram as well and just thinking of the road to travel in order to get to the "to prove" statement. But my kids just seem to flounder.. and that of course leads to their giving up before starting. Many kind thanks for being on the internet and being a resource. Sincerely, Mrs Buck cathbuck@earthlink.net

    ps I am in the city of El Segundo, in CA.

    1. Hi Cathy. Check out my more recent post on what I did this year. I have always left implicit the idea that proofs are a series of if-then statements that connect together, but this year I directly told them this during our first lesson on actually writing proofs. Check out how I did this at my recent post here: http://evanweinberg.com/2012/09/27/first-day-of-geometry-proofs-refining-my-process/

      In the beginning I've found that students are helped by being given the reasons and filling in the justifications. This is good for becoming acquainted for properties of equality, congruence, and definitions. without students having to construct the entire structure on their own. As soon as possible though, I try to get them out of this habit, because the coolest proofs don't follow a rigid set of procedures. The sooner you can have them writing proofs and getting feedback on them, the better. I don't know about specific books - I always pick and choose from different ones. Thank you for visiting - I'm eager to hear your ideas as well!

Leave a Reply