11 Problem Dashboard Functions
11.1 Assessment Functions
11.1.1 extract_assessment_json
11.1.1.1 Main Documentation
Convert a JSON object into a tidyJSON dataframe
Description:
Convert a JSON object into a tidyJSON dataframe
Usage:
extract_assessment_json(assessment_json)
Arguments:
assessment_json: An JSON file generated through `xml_extraction
<course> -assessments`
Value:
A flattened dataframe with the columns `url_name`, `title`,
`label`, `name`11.1.2 extract_assessment_csv
11.1.2.1 Main Documentation
Convert a CSV respresenting an open_assessment.sql query into a usable
format
Description:
Extracts the nececessary information from event JSON and discards
it. Points possible should always be the same for each assessment.
Usage:
extract_assessment_csv(assessment_tbl)
Arguments:
assessment_tbl:
Value:
An extracted table
Examples:
extract_assessment_csv(raw_assessment)
11.1.3 join_extracted_assessment_data
11.1.3.1 Main Documentation
Join the results of extract_assessment_csv and extract_assessment_json
Description:
This function joins the BigQuery and XML data. This is needed to
populate the BigQuery data with the title and label fields. If the
ID of the assessment does not occur in the XML, it is removed
before preceding.
Usage:
join_extracted_assessment_data(extracted_csv, extracted_json)
Arguments:
extracted_csv: the result of extract_assessment_csv
extracted_json: the result of extract_assessment_json
Value:
A joined dataframe of the two incoming dataframes
Examples:
join_extracted_assessment(sample_extracted_assessment_tbl, extracted_content)
11.1.4 summarise_joined_assessment_data
11.1.4.1 Main Documentation
Summarise assessment data for plotting
Description:
This function checks that the number of points possible does not
vary within title-label groups.
Usage:
summarise_joined_assessment_data(joint_assessment, trunc_length = 20)
Arguments:
joint_assessment: The result of join_extracted_assessment_data
trunc_length: The length of that the label should be truncated to. Do
not set to less than 4
Value:
A summarised dataframe of the average scores in each area.
Examples:
summarise_joined_assessment_data(sample_join_extracted_assessment_data, 20)
11.1.5 plot_assessment
11.1.5.1 Main Documentation
Plot the summary assessment data
Description:
Plot the summary assessment data
Usage:
plot_assessment(summary_assessment)
Arguments:
summary_assessment: the resulting dataframe from `summary_assessment`
Value:
A faceted bar plot11.2 Problems Functions
11.2.1 clean_multiple_choice
11.2.1.1 Main Documentation
Clean a demographic multiple choice CSV
Description:
This function cleans a CSV retrieved from the
`demographic_multiple_choice` SQL script. It transforms the
`sum_dt` column into activity_level. It also removes any
non-multiple choice problems.
Usage:
clean_multiple_choice(raw_csv)
Arguments:
raw_csv: A dataframe from the read_csv
Value:
A dataframe with the same dimensions as the CSV11.2.2 create_question_lookup_from_json
11.2.2.1 Main Documentation
Parse a JSON object and return it as a flat dataframe.
Description:
This function is required to convert the JSON derived from the
xbundle XML into a format that other dataframes can interact with.
Usage:
create_question_lookup_from_json(name_lookup)
Arguments:
name_lookup: A JSON object contain keys for `id`, `problem`,
`chapter_name`, `chapter_name`, `choices`, `correct_id` and
`correct`
Value:
A flat dataframe with the above columns11.2.3 filter_valid_questions
11.2.3.1 Main Documentation
Filter out invalid multiple choice problems
Description:
Removes problems that were marked as 100 problems where every
answer is correct). Also checks from the dataframe derived from
xbundle to ensure that the problem exists in the course.
Usage:
filter_valid_questions(problems_tbl, lookup_table)
Arguments:
problems_tbl: The result of `read_multiple_choice_csv`
lookup_table: The result of `create_question_lookup_from_json`
Value:
The filtered problems_tbl11.2.4 get_mean_scores
11.2.4.1 Main Documentation
Get mean scores.
Description:
This function is not affected by filters. If you need a function
that is, use `get_mean_scores_filterable`. Removes scores that do
not appear in the xbundle XML or have a 100 questions masquerading
as multiple choice problems.
Usage:
get_mean_scores(problems_tbl, lookup_table)
Arguments:
problems_tbl: The result of `clean_multiple_choice`
lookup_table: The result of `create_question_lookup_from_json`
Value:
A dataframe with each problem, the number of users that attempted
the problem and the percent score11.2.5 tally_correct_answers
11.2.5.1 Main Documentation
Count the number of correct answers for each question.
Description:
Count the number of correct answers for each question.
Usage:
tally_correct_answers(joined_scores)
Arguments:
joined_scores: `problems_tbl` and `lookup_table` joined by `id` and
`item_response`
Value:
A summarised dataframe of correct counts11.2.6 calculate_percent_correct_tbl
11.2.6.1 Main Documentation
Calculate the percent of correct answers
Description:
Each row represents a problem.
Usage:
calculate_percent_correct_tbl(joined_scores, number_correct)
Arguments:
joined_scores: `problems_tbl` and `lookup_table` joined by `id` and
`item_response`
number_correct: the result of `tally_correct_answer`
Value:
A dataframe where each row is a row and the columns show the
percentage and absolute number of students that got the problem
correct.11.2.7 get_mean_scores_filterable
11.2.7.1 Main Documentation
A filterable version of get_mean_scores
Description:
A filterable version of get_mean_scores
Usage:
get_mean_scores_filterable(filtered_problems_tbl)
Arguments:
filtered_problems_tbl: The result of `read_multiple_choice_csv`
filtered by demographics and chapter
name_lookup: A JSON object contain keys for `id`, `problem`,
`chapter_name`, `chapter_name`, `choices`, `correct_id` and
`correct`
Value:
A dataframe where each row is a row and the columns show the
percentage and absolute number of students that got the question
correct.11.2.8 join_summary_lookup
11.2.8.1 Main Documentation
Join the summary and lookup tables
Description:
Join the summary and lookup tables
Usage:
join_summary_lookup(summarised_scores, lookup_table)
Arguments:
summarised_scores: the result of `get_mean_scores_filterable`
lookup_table: The result of `create_question_lookup_from_json`
Value:
The inner join of the two dataframes11.2.9 summarise_scores_by_chapter
11.2.9.1 Main Documentation
Summarise scores by chapter
Description:
Calculates the average score on problems for each of the chapters
based.
Usage:
summarise_scores_by_chapter(summarised_scores, lookup_table)
Arguments:
summarised_scores: the result of `get_mean_scores_filterable`
lookup_table: The result of `create_question_lookup_from_json`
Value:
A dataframe where each row is a chapter that contains the average
score on that chapter11.2.10 prepare_filterable_problems
11.2.10.1 Main Documentation
Produce a dataframe clean dataframe of all questions
Description:
Returns the question name and it's rounded percentage (out of
100). This function is used to prepare the data table shown at the
bottom of the problem dashboard.
Usage:
prepare_filterable_problems(summarised_scores, lookup_table)
Arguments:
summarised_scores: the result of `get_mean_scores_filterable`
lookup_table: The result of `create_question_lookup_from_json`
Value:
A dataframe with the question and percent correct.11.2.11 get_extreme_summarised_scores
11.2.11.1 Main Documentation
Select the easiest (or hardest) problems
Description:
Set the index negative to receive the hardest problems.
Usage:
get_extreme_summarised_scores(summarised_scores, index)
Arguments:
summarised_scores: the result of `get_mean_scores_filterable`
index: the number of questions you wish to view
Value:
The summarised scores dataframe with the with the number of rows
equal to the absolute value of the index.11.2.12 join_problems_to_lookup
11.2.12.1 Main Documentation
Left join extracted problems and lookup table
Description:
Left join extracted problems and lookup table
Usage:
join_problems_to_lookup(extracted_problems, lookup_table)
Arguments:
extracted_problems: the result of `clean_multiple_choice`
lookup_table: The result of `create_question_lookup_from_json`
Value:
The left joined dataframe by id and choice_id, filtering out blank
problems11.2.13 join_users_problems
11.2.13.1 Main Documentation
The left join of joined problems and summarised scores
Description:
The left join of joined problems and summarised scores
Usage:
join_users_problems(joined_problems, summarised_scores)
Arguments:
joined_problems: `problems_tbl` and `lookup_table` joined by `id` and
`item_response`
summarised_scores: the result of `get_mean_scores_filterable`
Value:
A left joined dataframe of joined problems and summarised scores11.2.14 filter_counts
11.2.14.1 Main Documentation
Count the number of filtered users for each problem
Description:
Count the number of filtered users for each problem
Usage:
filter_counts(joined_user_problems)
Arguments:
joined_user_problems: The result of `join_users_problems`
Value:
A dataframe with the count of unique users for each question11.2.15 filter_extreme_problem_choices
11.2.15.1 Main Documentation
Retrieve the question and choice information from the extreme problems
Description:
Retrieve the question and choice information from the extreme
problems
Usage:
filter_extreme_problem_choices(lookup_table, extreme_problems, filtered_counts)
Arguments:
lookup_table: The result of `create_question_lookup_from_json`
extreme_problems: The result of `get_extreme_summarised_score`
filtered_counts: The result of `filter_counts`
Value:
A dataframe with lookup information attached11.2.16 aggregate_extracted_problems
11.2.16.1 Main Documentation
Perform the last wrangling before plotting extreme problems
Description:
Calculates the percent of students selected each option and
determines if that option is correct.
Usage:
aggregate_extracted_problems(joined_user_problems, extreme_problems,
question_choices)
Arguments:
joined_user_problems: The result of `join_users_problems`
extreme_problems: The result of `get_extreme_summarised_score`
question_choices: The result of `filter_extreme_problem_choices`
Value:
A dataframe with `problem,` `choice` and percent `correct`11.2.17 aggregate_melted_problems
11.2.17.1 Main Documentation
Aggregate questions before plotting
Description:
Aggregate questions before plotting
Usage:
aggregate_melted_problems(lookup_table, joined_user_problems, extreme_problems)
Arguments:
lookup_table: The result of `create_question_lookup_from_json`
joined_user_problems: The result of `join_users_problems`
extreme_problems: The result of `get_extreme_summarised_score`
Value:
A dataframe ready to be plotted by `plot_aggregated_problems`11.2.18 plot_aggregated_problems
11.2.18.1 Main Documentation
Plot aggregated problems
Description:
This funciton is used to plot the top or bottom questions.
Usage:
plot_aggregated_problems(agg_melted_problems)
Arguments:
agg_melted_problems: A dataframe with the problem and choice names as
well as the number of students who chose each option.
Value:
A facetted ggplot bar chart11.2.19 plot_problem_chapter_summaries
11.2.19.1 Main Documentation
Overview plot (Chapter by Course)
Description:
Overview plot (Chapter by Course)
Usage:
plot_problem_chapter_summaries(chapter_summary_tbl)
Arguments:
chapter_summary_tbl: The aggregated data on a per chapter basis
Value:
A ggplot bar chart