In this section, students examine three study types: experimental, observational, and survey. Students classify descriptions of studies as one of the three types and notice some aspects of study design that make them better at addressing a question of interest. One of the main aspects of good study design is the inclusion of randomness in the selection of participants in the study.

In this section, students focus on the normal distribution as a model for bell-shaped distributions. By using this model and technology (or tables of values), students can approximate the proportion of data from a distribution that is within certain intervals. This technique will be useful in later sections when students quantify expected confidence in estimates for population characteristics.

The first lesson of this section is an optional review of some statistical measures, such as mean and standard deviation, and of distribution descriptions preparing students to recognize distributions that would fit with a normal distribution model.

Histogram from 16.5 to 25.5 by 0 point 5’s. Handspan, centimeters. Beginning at 16.5 up to but not including 17, height of bar at each interval is .001, .007, .013, .016, .052, .059, .087, .132, .159, .132, .104, .098, .069, .032, .023, .010, .004, .002, 0. A bell-curve line is drawn at the top of each of the bars across the graph.

In this section, students see that samples can be used to estimate characteristics for a population and how to adjust for the fact that different samples would result in different estimates. First, students recognize that working with real data often means dealing with some amount of variability. For example, if a random sample suggests that half of a population has a trait, it might not be surprising if it is actually closer to 52% that have the trait. It might be surprising if 80% of the population does, though.

Students are introduced to the idea of a margin of error that can be attached to point estimates to indicate a range of values that would not be surprising based on the information from a sample. Then students practice estimating either a proportion or mean for a population using a sample, and including a margin of error for the point estimate based on variability from simulated samples.

In this section, students examine the difference in means between two groups in an experimental study and determine if there is enough evidence to determine if the difference is likely due to the treatment. Given data from two groups in an experiment, students create a randomization distribution and use what they know about normal distributions to recognize when the observed difference in means would be unusual if there was no effect of the treatment, and thus, when the treatment that defines the group likely causes some effect on the response variable.

Finally, students are given the chance to design their own experimental study, collect data, and analyze it to draw a conclusion about the experiment.

A note on the language of this section: The analysis happens under the assumption that the treatment of the experiment has no effect. This means that any conclusion that can be made from the analysis either supports that assumption or does not. At this level, it is okay for students to confuse a statement like “The analysis shows that there is not enough evidence to support the idea that the treatment has no effect” with a statement like “The analysis shows evidence that the treatment has an effect,” but use the opportunity to explain the difference to students.

In this final section, students have the opportunity to apply their thinking from throughout the unit. As this is a short section followed by an End-of-Unit Assessment, there are no section goals or checkpoint questions.