The goal of this exercise is to show students that several different models can be created from one set of observable data and that each model is equally acceptable if it predicts the observed results.

Students design a simple pencil and paper model to describe what might be happening inside the black box device. They construct their model of what is happening inside the tube from data gathered from outside the tube. Students then present their ideas to their peers.

MATERIALS
black box device (see Appendix A for building instructions)

METHOD

01. Arrange the black box so one of the top cords is extended. Pull the other top cord across so that it is now extended. Pull the top two cords back and forth a few times, ensuring that all students can see the apparatus. Each time you pull one cord, the other will retract into the black box. This will give the students the sense that the top two cords are, in fact, only one cord. Then pull one of the bottom two cords. Students will be surprised to see that pulling a bottom cord also causes a top cord to retract. Continue the demonstration by randomly pulling each cord. Ask students to call out predictions as you pull on the cords.

02. Allow students to try their own combinations, noting the motion and tension of the cords or anything else that might help them decipher how the cords are attached.

03. Now instruct students to complete this sketch by drawing their interpretation of how the cords might be attached. Make sure students do this individually.



04. Have several students share their sketches on the board.

05. Systematically test the accuracy of each student’s idea, analyzing the diagram to see if it could predict the behaviour that is actually witnessed when the cords are pulled. Each drawing will likely have at least a chance of working. This demonstration is most effective when there are at least 10 different drawings on the board, each with the potential of being the correct depiction of how the cords are actually attached. Refine the models drawn on the board by drawing out new predictions (e.g., Would the black box make a noise if it is shaken?).

Note: Never divulge how the black box device is actually connected. The models must be judged primarily on their ability to explain and predict the observations.

DISCUSION
Scientists use models to represent or simplify complex realities. Sometimes the models are so good at representing the reality that we forget they are models. Sometimes the reality is so complex that a simple physical model is inadequate.

One of the challenges arising from quantum physics is creating simple models that make sense in the classical world while remaining true to the reality of the quantum world. Wave–particle duality is an example of this challenge. There are no classical analogues that can accurately represent the behaviour of a quantum object, so we are left with a model that does not really make sense in the classical world.

Another challenge arising from quantum physics is that a model is measured primarily by its ability to explain the observations, and observing quantum systems can be problematic. There are real limits to what can and cannot be measured in the quantum world and, therefore, a limit to how refined the models can be. In the video we explore the various models, or interpretations, for what is happening during the double-slit experiment. Some interpretations may be preferred over others because they provide useful insights or make different assumptions, but each of them provides a complete description of the observed data. Any attempt to observe what the object is doing during the double-slit experiment alters the data and prevents us from refining our models.

SUGESTED USES

Pre-video: Use the activity to introduce a discussion about how physicists construct models. Draw out the concept that models are built in response to observations and should have predictive power.

Post-video: Use the activity to revisit the part of the video where scientists provide alternative interpretations for the double-slit experiment. Build on the analogy that the inside of the black box is like the “inside” of the double-slit experiment. We cannot know what is “actually” happening inside, so any model that successfully explains the observed data can be considered valid.