Track 3 · Visualising and Exploring Data · Lesson 1 of 11

Choosing the right chart

Choose a chart from the question you are asking and the types of the variables involved, and explain why other charts would mislead or hide the answer.

25 minFoundation9 graded exercisesFree

An HR analyst is asked three questions in one meeting: how staff are spread across departments, how salaries are distributed, and whether experience pays. Each question needs a different chart — and the wrong one can hide the answer or invent a pattern that is not there.

A chart is an answer to a question. Before drawing anything, ask two things:

What is the question? Comparing amounts between categories, describing how values are distributed, comparing distributions between groups, looking for a relationship, or following change over time.
What types are the variables? A categorical variable puts each case in a group (department, travel mode); an ordinal one has groups with a natural order (job level L1–L5); a numeric variable is a measured quantity (salary, minutes). Time is numeric but special: it has a direction, so we read it left to right.

The chart follows from those two answers far more often than from taste.

Question	Variables	Default chart	Also good
How many in each group?	One categorical	Bar chart (sorted unless the categories are ordered)	Dot plot of counts; a table for a few numbers
How are the values spread?	One numeric	Histogram	Box plot, density curve, ECDF
Do groups differ?	Numeric by categorical	Box plots side by side	Violin or strip plots; small-multiple histograms
Are two quantities related?	Two numeric	Scatter plot	Scatter with a fitted line; hexbin when points overlap
How did it change?	Numeric over time	Line chart	Small multiples for many series
How do two categorical variables combine?	Two categorical	Grouped or 100% stacked bars	A two-way table, a heatmap of counts

Question × variable types → a good default chart

Here is the first question applied to the (synthetic) salaries dataset: 600 employees in six departments. Department is categorical, so we count each group and draw one bar per department. Bar length encodes the count, and lengths starting from a common zero baseline are what the eye compares most accurately.

Employees per department. Operations is the largest (183 of 600); Human Resources is the smallest (41).

Show the data as a table

Department	Count
Operations	183
Sales	129
Technology	125
Finance	71
Human Resources	41
Legal	51

The second question — how are salaries distributed? — involves one numeric variable. A bar per distinct salary would be useless (almost every salary is unique), so a histogram groups the values into intervals (bins) and shows how many fall in each.

Monthly base salaries of 600 employees. Most are between about 7 000 and 20 000 SAR, with a long tail of senior salaries to the right.

Show the data as a table

From	To	Count
6630	8821.15	97
8821.15	11012.3	95
11012.3	13203.5	110
13203.5	15394.6	73
15394.6	17585.8	46
17585.8	19776.9	42
19776.9	21968.1	29
21968.1	24159.2	15
24159.2	26350.4	12
26350.4	28541.5	19
28541.5	30732.7	18
30732.7	32923.8	8
32923.8	35115	3
35115	37306.2	2
37306.2	39497.3	3
39497.3	41688.5	1
41688.5	43879.6	2
43879.6	46070.8	9
46070.8	48261.9	5
48261.9	50453.1	4
50453.1	52644.2	0
52644.2	54835.4	2
54835.4	57026.5	1
57026.5	59217.7	1
59217.7	61408.8	2
61408.8	63600	1

To compare salaries between job levels (numeric by ordinal), put one box plot per level on a shared axis. Each box spans the middle half of the salaries in that level and the line inside is the median. Keep the levels in their natural order, L1 to L5, so the staircase is visible. The median rises from 8,690 SAR at L1 to 46,020 SAR at L5.

Base salary by job level, in level order. Each level's box sits higher than the one before, and the boxes barely overlap.

Show the data as a table

Group	n	Lower whisker	Q1	Median	Q3	Upper whisker	Outliers	Mean
L1 Associate	183	6630	8060	8690	9600	11660	1	8859.89
L2 Specialist	195	9540	11790	12800	13725	16610	1	12864.6
L3 Senior Specialist	127	14410	16955	18820	20435	25060	1	18857.2
L4 Manager	62	22370	26970	28415	30112.5	34370	1	28398.4
L5 Director	33	37750	44050	46020	49110	55440	6	47056.7

The question does experience pay? links two numeric variables, so each employee becomes one point on a scatter plot: years of experience across, salary up. The cloud rises from left to right (Pearson's $r = 0.76$ ), and it fans out: salaries vary more among experienced staff.

Experience against base salary for 600 employees. Salaries tend to rise with experience, and their spread widens.

Show the data as a table

Experience (years)	Monthly base salary (SAR)
17.5	20260
2.4	9080
11	13000
12.7	12460
1.8	7740
16.8	16940
12.3	20110
5.3	14220
7.5	22110
1.5	8590
2	11560
6.3	9740
4.5	9710
7.8	10880
6.9	10980
7.9	12800
12.9	21570
7.9	8130
12.3	20690
4.6	24160
3.1	13500
10.6	10960
6.4	10180
18.4	38710
5.7	14360
13.1	11960
25.4	42390
8.9	12540
14.6	9090
11	17630
2.5	18040
2.5	9370
5.7	8560
7.2	13780
2.7	10120
14.9	23620
1.4	8890
27.7	30120
13.9	22370
5.7	11960
14.4	30090
12.9	27840
13.9	11950
2.4	8060
15.8	12120
24.9	63600
0.8	8380
6	27720
5.7	11300
12.6	15390
10.2	13960
8.8	11660
8.2	17820
19.3	17000
11.5	17170
16.9	19030
3.3	7980
9.2	11380
11.9	10140
5.7	12740
4.8	11550
15.6	29350
7.3	13140
22.2	44870
8.5	20930
6.4	14350
1.2	7270
6	10180
7.7	10620
11.5	16880
16.1	29440
18.2	22610
2.2	8360
7.9	12770
14	25820
12.6	30780
4.2	11170
13.4	18730
6.7	7090
12.3	18320
10.5	15520
8	11990
2.2	13440
1	9490
7.1	15650
8.2	20480
10.2	14820
10.3	19720
1.9	8290
6.7	12210
8.9	13010
10.6	13250
5.7	12310
5.2	14760
7.9	13660
11.2	21860
12.8	17230
11.1	13900
8.4	11450
2.6	9440
10.4	14440
5.2	10960
12.5	47310
2.7	10590
6.7	13080
8.8	11790
3.1	9220
7.4	10300
6.7	11410
9	16200
18.6	46590
31.3	44050
5.8	11230
7.2	8040
7.6	11100
5.3	9850
1.8	9240
10.8	15740
7.4	9550
4.9	8580
4.1	7780
1.4	7910
3.8	8350
9	15100
2.7	7950
5.7	10920
8	9220
14.7	13150
6.3	8570
4.6	9400
11	19100
12.3	22040
3.6	7820
17.1	12630
16.3	14790
11.5	20840
5.4	9800
6.4	8330
11.6	14520
5.3	10550
13.1	16250
11.1	12840
7.9	15260
15.2	18490
11.2	13140
3.9	8460
1.6	18680
2.3	9280
8.2	8300
5.3	12380
21.3	18980
20.6	32840
6.9	12690
4.4	11490
4.3	8090
7.3	14400
11.2	16860
13.3	20540
5.3	12120
22.5	28400
5.4	9050
3.6	9190
3.4	8740
15.3	13620
13.8	19150
11	12930
9.6	11830
8.7	14150
9.3	14310
7	12830
7.3	14240
13.4	13450
9.3	19850
12.4	13760
8.6	11170
24.9	22540
1.8	8270
6.6	7850
11.9	15500
16.5	22050
19.4	19280
25	44160
1	9630
5.7	10900
10.4	17940
17.5	19730
8.2	18820
3.6	7720
25.6	29020
10.4	22170
1.8	8900
10.8	20490
7.4	12930
5.2	13310
10.4	10120
8.1	7780
1.9	15420
8.8	13690
23.9	46870
8.9	8500
9.2	13540
4	9810
11.1	13800
25.1	29680
22.5	34370
35.5	61210
2.5	7530
4.8	14470
4.8	12300
11.6	11900
5.3	7950
4.9	9000
25.9	49110
11.5	26610
2.2	19610
2.2	10480
10.5	13240
9.6	8950
9.7	16560
2.4	8110
15.5	15790
8	10630
3.8	10860
10.1	11330
5.1	13510
13.9	15830
8.7	16850
9.6	12830
3	7640
10.2	20120
5.8	9180
16.7	54000
6.3	8210
8.5	13250
3	13420
5.7	14020
12.3	15400
8.4	25590
14.8	16790
23	19280
17.2	11830
4.1	7790
12.2	28590
13.5	33740
19.9	32390
4.8	10400
6.8	8030
4.6	7850
3.6	12800
9.6	12190
22.7	27730
9.7	29450
5.2	12450
5	19550
11.9	25440
9.3	11380
11.5	16880
11.3	14370
11.4	10340
4.5	12610
2.3	10370
4.2	11950
5.2	7380
10.7	10850
17	19470
20.3	17790
6.5	8300
9.7	10410
10.2	12680
34.4	55440
4.3	9930
4.5	15160
11.3	11980
4.8	8880
9.6	8330
26.6	54590
5.5	11910
12	27560
5	13490
15.8	19370
6.3	8430
5.6	7720
6.5	8100
9	15460
3.9	8950
13.1	12020
8.2	11860
1.6	13090
12.3	13260
14	8720
7.5	25330
1.8	8410
11.5	11610
10.2	12980
8.8	14150
12.8	29280
18.6	23800
3.2	8060
12	11350
4.5	8320
13.3	16410
17.2	16610
4.4	16900
21.3	12690
7.7	10080
6.1	17000
12.3	21610
10.9	17430
9.2	7550
10.8	19300
5.3	7110
6	9600
0.9	7980
2.8	8730
4.8	9530
15	27260
15.7	21000
5.9	8340
3	10860
3.1	18170
17.9	17260
3.4	13840
11.2	12610
10.5	20940
7.3	12610
5.9	12000
0.5	8980
9.5	16740
6.7	9010
9.5	25670
15.6	28960
8.4	12690
30.2	35240
5.3	9860
3.6	8300
3.9	9000
19.9	19590
23.9	19810
2.8	9220
17.5	16950
14.1	27720
24.5	15450
12.2	12970
33.5	48310
5.2	11290
7.1	7450
15.7	21510
19.4	20560
13.4	28290
15.7	28980
17.4	23710
19	19730
3.2	14410
14	21780
3.2	7790
10	28370
10.4	11060
7.5	12200
9.9	11080
2.6	11270
15.6	16920
5.4	13590
15.6	16540
12.9	11510
11	20160
0.8	11380
4.1	10340
16.2	19400
3.1	11720
16.3	28430
3.8	9340
7.5	12370
5.1	9570
9	15840
7.9	19930
18.1	31690
9	13250
17.3	30960
6.2	8060
7.8	12830
6.7	12070
5.4	8540
3.7	9150
6.3	12590
16.8	17140
34.2	48520
7.9	13760
9.4	13430
9.1	13540
8.5	10990
7.3	9350
19.8	42320
21.6	28460
37.7	47070
4.9	16770
7.8	30470
15.1	16960
22.1	18970
5.8	15380
10.4	9490
9.3	9270
11.5	8130
23.5	27910
38	44700
2.4	9550
13.7	14950
6	8460
20.6	46860
2.8	8690
2.7	10460
17.1	21280
26.1	44740
9.1	21120
5.3	17750
7.7	14160
10.8	17910
13.9	28030
7.2	9930
19.7	27180
18	19630
3.2	11210
13.6	40550
15.9	23530
13.2	17460
3.5	8200
14.2	28880
10	21760
11.9	46020
7.2	9780
5.9	12550
7.8	11900
3.7	25410
4.8	16930
37	60430
2.5	8720
5.2	13280
24.7	44740
16.7	17970
11	14320
5.2	13350
4.9	17360
10.7	27750
15.8	15110
18.3	31270
13.8	18930
10.2	8140
8	8550
22.9	18970
14.7	27790
7.5	13290
13.1	15130
7.8	8150
7	11630
8.2	7780
13.3	12820
12.8	19610
10.4	15370
13.6	30330
2.9	10740
2.4	8400
6.7	11390
3.8	8460
22	45820
2.8	10120
6.6	7130
1.9	7540
7.2	8260
9.7	11790
9.9	14980
2.7	7590
4.6	9460
21.1	39460
25	18810
9	10800
2.9	8740
17.3	12870
2.4	7400
6	11230
13.2	25860
6.2	11150
4.5	18180
7.8	15720
8.7	15350
7.8	30120
7	9600
4.6	7980
8.3	11700
4.3	9830
9.6	14540
2	11560
16.4	27120
17.6	30370
16.6	13340
2.6	11130
18.2	25060
16	10360
20.5	23310
19.6	20390
8.9	14040
7.7	7590

First 500 of 600 rows.

Finally, numbers recorded over time call for a line chart, because the line joining consecutive months shows direction and rhythm. Below, eight years of monthly demand for three products: each has its own seasonal peak.

Ceiling fan
Water heater
School backpack

Monthly units ordered for three products over 96 months (synthetic). Fans peak every summer, water heaters every winter, backpacks every August.

Show the data as a table

Month index	Ceiling fan	Water heater	School backpack
1	1232	2430	1080
2	1481	2264	429
3	1948	1483	453
4	2427	1304	539
5	3360	876	504
6	4211	837	489
7	4240	629	944
8	4524	726	3144
9	2822	937	1183
10	2242	1382	495
11	1584	2008	534
12	1224	2611	462
13	1440	2616	1091
14	1385	2242	551
15	1761	1566	410
16	2551	1282	448
17	3545	1022	508
18	4415	884	531
19	4039	747	1054
20	4351	792	3144
21	3335	1006	1105
22	2247	1455	586
23	1472	1923	534
24	1233	2317	520
25	1301	2762	1070
26	1481	2234	578
27	1957	1634	432
28	2813	1211	410
29	3940	852	485
30	4544	773	575
31	4511	674	988
32	4159	746	3616
33	3133	937	1371
34	2171	1476	529
35	1532	2119	527
36	1641	2863	470
37	1299	2635	1154
38	1412	2391	523
39	2004	1891	372
40	2991	1263	429
41	3951	1051	526
42	4086	762	514
43	4887	690	1030
44	4262	790	3049
45	3731	1103	1157
46	2230	1625	563
47	1735	2193	462
48	1514	3027	486
49	1362	2629	975
50	1891	2255	542
51	2167	2020	431
52	3136	1277	445
53	4405	959	605
54	4744	916	470
55	5290	615	974
56	4815	842	3317
57	3244	1071	1215
58	2415	1611	514
59	1690	2000	511
60	1232	2637	491
61	1535	1805	1167
62	1768	1898	553
63	2208	1326	432
64	2791	862	481
65	4240	816	440
66	5441	652	549
67	4976	532	1000
68	4866	644	3237
69	3403	873	1266
70	2631	1116	566
71	1961	1720	555
72	1310	2102	454
73	1627	1871	1184
74	1985	1943	531
75	2508	1514	508
76	3386	962	433
77	3874	834	511
78	5516	633	502
79	5929	659	1025
80	4521	650	3207
81	4182	870	1173
82	2768	1213	629
83	2024	1507	526
84	1399	2115	487
85	1561	2291	1061
86	1766	1721	473
87	2401	1296	501
88	2751	1052	547
89	4458	779	525
90	5186	724	557
91	5489	578	992
92	5730	702	2938
93	4069	872	1256
94	2937	1233	460
95	1881	1780	465
96	1731	2036	474

Worked example

Choosing for a manager's question

A manager asks: “Do Sales staff get bigger bonuses than Technology staff?” Which chart should you draw from the salaries data?

Question type. It compares a quantity between groups.
Variable types. annual_bonus_sar is numeric; department is categorical (we keep just two of its levels).
Default chart. Numeric by categorical → box plots side by side on one shared axis.
Check the data. About 29% of bonuses are exactly zero. A box plot alone can hide that spike, so add the share with no bonus in the caption, or show strip plots of the individual values.

Side-by-side box (or strip) plots of bonus for Sales and Technology on a common axis, with the percentage receiving no bonus reported for each.

Quick check

You want to show what share of 600 employees work in each of six departments. Which chart is the safest default?

Exercises

The lesson is complete when every exercise is answered correctly. Hints are free; viewing a worked solution before solving means the exercise won't count towards mastery.

0 of 9 exercises correct (not saved — sign in to keep your progress)

Exercise 1 · Guided, difficulty 1 of 3

From question to chart

Not started

Sort each question into the chart that answers it best by default.

Which payment method do customers use most often?

How are basket values (SAR) spread — symmetric or with a long tail?

Are flight delays longer for some airlines than others?

Do larger homes sell for higher prices?

How have weekly sales at one store changed over two years?

How many employees are in each job level, L1 to L5?

Is engagement score different across six departments?

Exercise 2 · Practice, difficulty 1 of 3

Match the variables to a chart

Not started

Each item names variables from the library datasets. Match each to the most suitable default chart.

Exercise 3 · Application, difficulty 1 of 3

Read the department bar chart

Not started

The bar chart in the lesson shows how many of the 600 employees work in each department. What percentage of employees work in Operations? Use the salaries dataset (or the chart's data table) and give the answer to 1 decimal place.

Data: salaries · Open in the lab to explore and calculate.

Exercise 4 · Application, difficulty 2 of 3

Read a median off the box plots

Not started

In the box plots of base salary by job level, read the median monthly base salary of L4 Manager employees. Then check it from the data (the dataset here is filtered to the L4 rows). Enter the median in SAR; a reading within 2% is accepted.

Data: salaries (job_level = L4 Manager) · Open in the lab to explore and calculate.

Exercise 5 · Interpretation, difficulty 2 of 3

A pie of averages

Not started

A colleague draws a pie chart with one slice per department, where each slice's size is the department's average base salary. What is wrong with this chart?

Exercise 6 · Practice, difficulty 2 of 3

Charts that can compare distributions

Not started

You want to compare the distribution of base salary across the five job levels (not just one average per level). Select all charts that can do this well.

Exercise 7 · Practice, difficulty 1 of 3

Steps in choosing a chart

Not started

Put these steps for choosing and checking a chart in a sensible order.

Exercise 8 · Practice, difficulty 1 of 3

How many employees in the biggest group?

Not started

A bar chart shows the number of employees in six departments of another (fictional) company:

Department	Operations	Sales	Technology	Finance	Human Resources	Legal
Employees	103	77	120	22	21	92

What percentage of all employees are in the department with the tallest bar? Give your answer to 1 decimal place.

Exercise 9 · Interpretation, difficulty 2 of 3

Justify your chart

Not started

A director asks: “Do employees with more experience earn more — and is that pattern the same in every department?” In two or three sentences, say which chart you would draw from the salaries data and justify it (what goes on each axis, and how you would show department).

What you can now do

Start from the question and the variable types, not from a favourite chart
Use bars for categories, histograms for one numeric variable, box plots for groups, scatter plots for two numeric variables and lines for time
Read counts, shares and medians accurately from bar charts and box plots
Explain why pies of averages and lines through categories mislead