Data Detectives: A Data Science Program for Middle Grade Learners

Figures & data

Fig. 1 Data Science Process Workflow for Data Detectives Program adapted from workflow from Grolemund and Wickham (2016).

Table 1 Data science practices for data detectives program adapted from Jiang and Kahn (2019).

Practices	Identifying data patterns	Talking about trends in data over time or across variables. Making sense of data.
	Pursuing data surprises	Observing an unexpected analysis in the data and making claims about it.
	Reasoning about data relationships	Making claims using data or explaining why a result occurred.
	Making data predictions	Making a guess or a prediction as to what the data will reveal after analysis.

Fig. 2 An example of the whole group Google trend activity discussion comparing pizza and tacos over a 7-day period.

Fig. 3 An example of a participant’s Google trend activity comparing four fast-food restaurants.

Fig. 4 Snack Data activity example of scatterplot and bar chart created with participants inputted data shared in spreadsheet.

Fig. 5 Example of prebuilt code in the R Jupyter notebooks for the Introduction to R lesson.

Fig. 6 Example of prebuilt code in the R Jupyter notebooks for the Pixar and Dreamworks lesson.

Table 2 Coding system used to evaluate participants work and reflective responses.

Category	Code	Definition
Process	Formulate Questions	Explicitly state formulate or come up with questions. Ask a statistical question.
	Collect Data	Explicitly state something about collecting data. Evidence of collecting, downloading, or searching for data.
	Data Analysis	Explicitly stating something about analyzing, modeling, or visualizing data. Statements around interpreting their data or making plots.
	Communicate Results	Explicitly stating something about communicating results. Statements around making claims about their data analyses.
Practices	Identifying data patterns	Talking about trends in data over time or across variables. Making sense of data.
	Pursuing data surprises	Observing an unexpected analysis in the data and making claims about it.
	Reasoning about data relationships	Making claims using data or explaining why a result occurred.
	Making data predictions	Making a guess or a prediction as to what the data will reveal after analysis.
Knowledge	Plots	Discussing how to choose among different plots and what the affordances are of each plot (histograms, scatterplots, tables, etc.)
	Correlations	Discussing how to use correlations to determine how similar two variables are.
	Effect Size	Identifying and discussing how effect size is useful to determine differences among datasets.

Fig. 7 Percentages of participants that engaged in data science practices and process in the Pixar and Dreamworks Activity.

Fig. 8 Percentages of participants that engaged in data science practices and process in their Final Project.

Fig. 9 Example of one Participant’s Final Project Scatterplot.

Fig. 10 Percentages of Participants that had a basic understanding of the statistics content covered.

Fig. 11 Pre- and post-surveys of participants attitudes toward data science.

Grolemund, G., and Wickham, H. (2016), “R For Data Science. OReilly Online Learning,” Available at https://www.oreilly.com/library/view/r-for-data/9781491910382/

 Google Scholar

Jiang, S., and Kahn, J. (2019), “Data Wrangling Practices and Process in Modeling Family Migration Narratives with Big Data Visualization Technologies,” in International Conference on Computer Supported Collaborative Learning.

 Google Scholar