STUDY GUIDE FOR EXAM 2
General tips
● The exam will have 25 questions + 2 extra credit questions. You will have 90 minutes.
● The extra credit questions at the end can help improve your score.
● This exam will be very much like the take-home assignments and Exam 1 in the style of
questions: There will be both MCQ/True-False and short answer questions. You could expect
roughly 10-12 short answer questions and 15-16 MCQ/T-F questions.
● The exam will be closed book, and it will be a pen-paper exam, not on Brightspace.
● You will be tested on content that is important for you to know in the long run, such as
logical/conceptual topics, or broad/important knowledge-based questions.
● There will be more questions from the slides/lectures than from the readings.
● Questions from the readings will be more about concepts conveyed in them rather than overly
specific info. You don’t need to memorize specific things from the readings (more info in the
table below)
● Several of you get things wrong because you didn’t read the question carefully. Please re-read
every question once you’re done to make sure your answer is an answer to the question asked.
● If you don’t know the answer to some question, that’s okay. I encourage you to provide your
best answer rather than leave it blank. Sometimes, the question is easier than it looks, and you
know more than you think.
● Ultimately, I’m putting this together because I want you all to do well in the exam. But this will
be useful only if you study!
WEEK CONTENT YOU NEED TO STUDY
5 Lecture:
1) Identifying features of graphs; different types of graphs including when and why we use
each of them; for each type of graph, you should learn to read/interpret and critique
graphs; what makes graphs good vs bad; for every graph in the slides, you should
practice interpreting it and evaluating whether it’s good vs bad/misleading
- Features of graphs:
- Title
- X-axis: values, label
- Y-axis: values, label
- Legend
- Data points, lines, bars
- Types of graphs:
- Pie charts:
- To show proportions of different categorical and mutually-exclusive
options
- Proportions need to add up to 1 (percentages need to add up to 100%)
- Bar plots:
- To compare values associated with separate (usually categorical
variables)
- Can be vertical or horizontal
- Can quickly become overwhelming if there’s too many categories
, - Bar plots showing averages should have error bars
- Important to include error bars if the height of the bar is an
average over several values
- Error bars show uncertainty or confidence in the average
- Wider bar = more variation/uncertainty
- Narrower bar = less uncertainty/more confidence
- Can sometimes show data points
- Depending on circumstance, may be better to plot individual data
points
- Line graphs:
- Only appropriate when x-values are continuous
- Frequently used for time-series data
- Can involve 2 y-axes (with caution)
- Can have error bars/bands
- Scatter plots:
- Show every data point
- Can help show relationships between variables while looking at all the
data
- Histograms:
- Bar plots that specifically plot frequency (ie., the number of times a certain
value occurs in the data)
- The x-axis has a variable and y-axis is always frequency
- Bin width: the intervals or ranges into which the data is grouped to create
the bars of the histogram
- Histograms vs. bar plots:
- Bar graph: has gaps, y-axis has some variable, x-axis has categories
- Histogram: no gaps, y-axis is frequencies, x-axis is number ranges
- Critiquing plots:
- Don’t invert the axes: upwards = increase
- axes/legends shouldn’t change midway
- Truncating axes can sometimes exaggerate relative differences
- Sometimes okay to truncate axes to a reasonable range for a measure
(especially when relative magnitude matters)
- Okay when you want to show change over time
- Use appropriate colors:
- Use basic elements of color for representation:
- Magnitude (low to high; 0-50) = light to dark shade of same color
(eg., light red to dark red)
- divergent/bidirectional (negative to positive; -15 to 15) = dark
shade of color 1 to dark shade of color 2 (eg., dark blue to white to
dark red)
- Categorical = different colors
- Consider audience and accessibility (red-green color blindness)
- Varying aspect ratio impacts interpretation
General tips
● The exam will have 25 questions + 2 extra credit questions. You will have 90 minutes.
● The extra credit questions at the end can help improve your score.
● This exam will be very much like the take-home assignments and Exam 1 in the style of
questions: There will be both MCQ/True-False and short answer questions. You could expect
roughly 10-12 short answer questions and 15-16 MCQ/T-F questions.
● The exam will be closed book, and it will be a pen-paper exam, not on Brightspace.
● You will be tested on content that is important for you to know in the long run, such as
logical/conceptual topics, or broad/important knowledge-based questions.
● There will be more questions from the slides/lectures than from the readings.
● Questions from the readings will be more about concepts conveyed in them rather than overly
specific info. You don’t need to memorize specific things from the readings (more info in the
table below)
● Several of you get things wrong because you didn’t read the question carefully. Please re-read
every question once you’re done to make sure your answer is an answer to the question asked.
● If you don’t know the answer to some question, that’s okay. I encourage you to provide your
best answer rather than leave it blank. Sometimes, the question is easier than it looks, and you
know more than you think.
● Ultimately, I’m putting this together because I want you all to do well in the exam. But this will
be useful only if you study!
WEEK CONTENT YOU NEED TO STUDY
5 Lecture:
1) Identifying features of graphs; different types of graphs including when and why we use
each of them; for each type of graph, you should learn to read/interpret and critique
graphs; what makes graphs good vs bad; for every graph in the slides, you should
practice interpreting it and evaluating whether it’s good vs bad/misleading
- Features of graphs:
- Title
- X-axis: values, label
- Y-axis: values, label
- Legend
- Data points, lines, bars
- Types of graphs:
- Pie charts:
- To show proportions of different categorical and mutually-exclusive
options
- Proportions need to add up to 1 (percentages need to add up to 100%)
- Bar plots:
- To compare values associated with separate (usually categorical
variables)
- Can be vertical or horizontal
- Can quickly become overwhelming if there’s too many categories
, - Bar plots showing averages should have error bars
- Important to include error bars if the height of the bar is an
average over several values
- Error bars show uncertainty or confidence in the average
- Wider bar = more variation/uncertainty
- Narrower bar = less uncertainty/more confidence
- Can sometimes show data points
- Depending on circumstance, may be better to plot individual data
points
- Line graphs:
- Only appropriate when x-values are continuous
- Frequently used for time-series data
- Can involve 2 y-axes (with caution)
- Can have error bars/bands
- Scatter plots:
- Show every data point
- Can help show relationships between variables while looking at all the
data
- Histograms:
- Bar plots that specifically plot frequency (ie., the number of times a certain
value occurs in the data)
- The x-axis has a variable and y-axis is always frequency
- Bin width: the intervals or ranges into which the data is grouped to create
the bars of the histogram
- Histograms vs. bar plots:
- Bar graph: has gaps, y-axis has some variable, x-axis has categories
- Histogram: no gaps, y-axis is frequencies, x-axis is number ranges
- Critiquing plots:
- Don’t invert the axes: upwards = increase
- axes/legends shouldn’t change midway
- Truncating axes can sometimes exaggerate relative differences
- Sometimes okay to truncate axes to a reasonable range for a measure
(especially when relative magnitude matters)
- Okay when you want to show change over time
- Use appropriate colors:
- Use basic elements of color for representation:
- Magnitude (low to high; 0-50) = light to dark shade of same color
(eg., light red to dark red)
- divergent/bidirectional (negative to positive; -15 to 15) = dark
shade of color 1 to dark shade of color 2 (eg., dark blue to white to
dark red)
- Categorical = different colors
- Consider audience and accessibility (red-green color blindness)
- Varying aspect ratio impacts interpretation
