Purpose of a data visualisation: exploratory data analysis

Name:

Date:

Class:

Quick Questions

Shade in the A or B circle for each question.

1. Who is the main audience of a visualisation made for exploratory data analysis?

A The analyst who is exploring the data

B A client who needs a polished final report

2. Which is the main purpose of exploratory data analysis?

A Discover patterns, relationships and outliers

B Present a final, polished message to the public

3. An analyst draws a scatter graph, then a histogram, then a column chart of the same data to see what each shows. What is this called?

A Choosing the one correct chart before starting

B Trying different charts to explore the data

4. Which chart is a good quick check for a possible relationship between hours studied and test score?

A A scatter graph of hours against score

B A pie chart of the hours studied

5. After a quick chart shows one point far from all the others, what should the analyst do next?

A Delete it straight away without checking

B Form a question, such as whether it is an error or a real value

Match the Pairs

Draw a line connecting each item on the left to its match on the right.

Left column		Right column
1=CORREL(B2:B9,C2:C9)		AExploring means trying several different charts and summaries to form questions. One chart is not enough to find patterns, relationships or outliers.
2Insert > Scatter graph (Maths vs English)		B=MAX(B2:B9)-MIN(B2:B9)
3Exploring: the analyst tries a quick summary of a column to see how the values are spread before deciding what to ask		CExploratory charts are quick and rough, and the analyst is the audience. Polishing them for others belongs to presenting, not exploring.
4Exploring: the analyst counts how many values fall in a group to look for a pattern or an unusual value		DTrying a quick chart to see patterns, relationships and outliers, with the analyst as the only audience
5Exploratory chart: polish colours and fonts for the final audience before looking at the data		EA quick summary check to see whether two columns are related, which can give an analyst a question to explore
6Make one chart, draw a conclusion and stop exploring		F=COUNTIF(B2:B9,">=80")

Definitions

exploratory data analysis

Fill in the Blanks

Complete each line of code by filling in the missing parts.

=______(B2:B9,C2:C9)

=___(B2:B9)-MIN(B2:B9)

SELECT ___(mark) FROM results;

SELECT * FROM results ORDER BY ____ DESC;

=COUNTIF(B2:B9,"_50")

Spot the Error

Shade in the letter of the line that contains an error.

Question 1

A =MIN(B2:B9)

B =MAX(B2:B9)

C =AVERAGE(B2:B9

Question 2

A Q: Does heat affect rain?

B =CORREL(B2:B8,D2:D9)

C Chart: scatter of Max, Rain

Question 3

A Step 1: Ask a question

B Step 2: Try a quick chart

C Step 3: Look for patterns

D Step 4: Publish final poster

Question 4

A Audience: the public

B Purpose: explore the data

C Output: quick draft charts

Question 5

A SELECT house, AVG(points)

B FROM students

C GROUP house;

Answer Key

Purpose of a data visualisation: exploratory data analysis

Section A: Quick Questions

A — The analyst who is exploring the data

A — Discover patterns, relationships and outliers

B — Trying different charts to explore the data

A — A scatter graph of hours against score

B — Form a question, such as whether it is an error or a real value

Section B: Match the Pairs

=CORREL(B2:B9,C2:C9) → A quick summary check to see whether two columns are related, which can give an analyst a question to explore

Insert > Scatter graph (Maths vs English) → Trying a quick chart to see patterns, relationships and outliers, with the analyst as the only audience

Exploring: the analyst tries a quick summary of a column to see how the values are spread before deciding what to ask → =MAX(B2:B9)-MIN(B2:B9)

Exploring: the analyst counts how many values fall in a group to look for a pattern or an unusual value → =COUNTIF(B2:B9,">=80")

Exploratory chart: polish colours and fonts for the final audience before looking at the data → Exploratory charts are quick and rough, and the analyst is the audience. Polishing them for others belongs to presenting, not exploring.

Make one chart, draw a conclusion and stop exploring → Exploring means trying several different charts and summaries to form questions. One chart is not enough to find patterns, relationships or outliers.

Section C: Definitions

exploratory data analysis — Using quick charts and summaries to discover patterns, relationships and outliers in data. The analyst is the audience, trying different charts to form questions.

Section D: Fill in the Blanks

=CORREL(B2:B9,C2:C9)

=MAX(B2:B9)-MIN(B2:B9)

SELECT AVG(mark) FROM results;

SELECT * FROM results ORDER BY mark DESC;

=COUNTIF(B2:B9,"<50")

Section E: Spot the Error

C — The closing bracket is missing, so the formula will not run. It should be =AVERAGE(B2:B9). Quick summaries like min, max and average are an early step in exploring data.

B — The two ranges are different sizes (B2:B8 and D2:D9), so CORREL gives an error. Both ranges must match: =CORREL(B2:B8,D2:D8). This tests the analyst's question about a relationship.

D — Exploratory analysis is a loop: what the analyst finds leads to new questions. Publishing a final poster is presentation, not exploration. The line should be 'Step 4: Ask a new question'.

A — In exploratory data analysis the analyst is the audience, because they are looking for patterns for themselves. The public is the audience for a presentation visualisation. It should be 'Audience: the analyst'.

C — The SQL keyword is GROUP BY, not GROUP. The correct line is GROUP BY house; and it summarises points for each house so the analyst can compare them.