We offer the following suggestions for implementing a data collection, longer term project in an introductory statistics course.

Suggestion #1: Let students take control of the project

Help the students take lots of ownership, such as developing their own project topic or allowing them to choose project topics from a list.  This suggestion can often outweigh the desire to have similar levels of difficulty among the projects.  Encourage students to submit a proposal that you can review and tweak. Encourage (perhaps with points) creativity in project topics. Randomized experiments are often easier to develop than truly random samples.  Consider incorporating a peer review of final project proposal. Don’t let students get discouraged by “insignificant” results.

Suggestion #2: Help students manage the project

Assign intermediate due dates with specific deliverables (these can be low stakes).  You can also encourage students to collect some pilot data before undertaking the full data collection plan.   Students may also want help in assigning “roles” to individuals or setting up a project group wiki in a course management system. Give very specific guidelines as to the final report (including do’s and don’ts) and rubric.  Encourage submission of a rough draft that you can provide quick feedback on.  Keep a close eye on overall workload.

Suggestion #3: Strongly suggest (require) collaborative efforts

Groups of 4 seem best for a more long-term project: If someone drops out, there is still a critical mass; if there is a division in the group, you can still have two pairs. More than four people tends to make scheduling meetings outside of class more difficult.  You may choose to build in a dependency structure (e.g., everyone gets the same grade) and make sure any final report they write comes across as “one voice.” 

Suggestion #4: Build in individual accountability

You can avoid many problems with group dynamics by letting them know early that their grades will be partly determined by the other students in their group.  You can ask students to grade each other and comment on the tasks completing by each person.  Students often take their peers more seriously and honestly. If there is a consensus, I do lower/raise individual grades. The individual component can be anywhere from 15-25%.  Other methods include putting a project-specific question on a course exam.

Suggestion #5: Also consider

- requiring an oral presentation and/or poster (great review/fodder for final exam)

- providing examples of previous topics and/or reports

- having a former student or group come into class to talk about their experience

- requiring a mixture of quantitative and qualitative variables

- collecting data across classes, institutions (check some IRB rules)

Back to Student Projects