Takeouts on/off

Rethinking Questionnaires charting (part 1)

 

Rethinking Questionnaires - Live data charting

Continuing our series on Rethinking Questionnaires, this demonstration is a preliminary warm-up to support our next discussion looking at using questionnaire techniques to develop the reporting process.

We're calling it a warm-up because, although we assume that most researchers use live data reporting - that is turning survey data into a chart set for analysis and presentation straight from the data, we do know some survey packages only create tables, and others only create charts one-by-one and not slide sets. The aim of this pre-demonstration is to establish a shared perspective on current reporting methods before we look more deeply at new ideas around reporting.

Here we're using YouGov data captured before in May 2016, just before the Brexit vote, as it has to extremely interesting response data which we will focus on next time. The charts are 'live' - that is they are driven directly by survey data. That means we can set a filter and update the whole chart set, or click to see drill-down charts and do things like sort and filter the data.

Click our logo for the next slide

(Filter: females)

dobney.com

dobney.com

 

The charting process

With live data charting, researchers have tools to manage respondent quality checks, data edits and open-ended coding. Charts then draw directly from original survey questions, but include summary lines, aggregated charts or other calculations, take-out text, and obviously text slides like this one.

Charting is web-based and can drive sophisticated charting tools like d3. Data can be exported to SPSS or R, and chart sets printed to .pdf or exported to Excel or Powerpoint. And tables can also be generated on the fly (see here).

As the data is live, we can add filters and have the whole dataset instantly updated. For instance, filter the data by Male or Female .

In practice, we normally present straight from the live slide set. If required the charts can be set up with a full animated Prezi style display.

Click our logo for the next slide

(Filter: females)

dobney.com

dobney.com

 

The data source

The dataset for this presentation is taken from YouGov data in connection with the UK Brexit vote. The data actually comes from May 2016 prior to the vote as part of the pre-referendum polling. 2969 respondents were interviewed across the UK by phone and online.

Opinion polling is struggling to predict the outcome of popular votes and elections, and we'll show a little more on this in the next demonstration. Here we'll show voting intention per party and for the referendum from the Yougov data, before looking at a couple of interesting questions about response rates.

The data we show are unweighted.

Click our logo for the chart on voting intention...

(Filter: females)

dobney.com

dobney.com

 

Voting intention May 2016

Base: 1525

All

Conservative
24
Labour
25
Liberal Democrat
5
UKIP
11
SNP
4
Plaid Cymru
1
Some other party
4
Would not vote
13
Don't know/refused
12
Base: 505

South

Conservative
27
Labour
20
Liberal Democrat
6
UKIP
14
SNP
0
Plaid Cymru
0
Some other party
6
Would not vote
15
Don't know/refused
12
Base: 152

London

Conservative
17
Labour
39
Liberal Democrat
5
UKIP
8
SNP
0
Plaid Cymru
0
Some other party
5
Would not vote
9
Don't know/refused
17
Base: 270

Midlands

Conservative
29
Labour
25
Liberal Democrat
2
UKIP
13
SNP
0
Plaid Cymru
1
Some other party
4
Would not vote
12
Don't know/refused
14
Base: 386

North

Conservative
21
Labour
32
Liberal Democrat
5
UKIP
11
SNP
0
Plaid Cymru
2
Some other party
3
Would not vote
15
Don't know/refused
11
Base: 134

Scotland

Conservative
19
Labour
15
Liberal Democrat
5
UKIP
3
SNP
43
Plaid Cymru
2
Some other party
1
Would not vote
6
Don't know/refused
6
Base: 78

Wales

Conservative
15
Labour
23
Liberal Democrat
5
UKIP
10
SNP
4
Plaid Cymru
5
Some other party
5
Would not vote
23
Don't know/refused
9
X The first of our charts is voting intention from May 2016. The main chart is the UK as a whole and regional charts are shown to the right and can be scrolled through. Green and purple bars show significant differences. The box with [ + ] gives options for the chart. Click the title of a subgroup chart to see it in more detail.

Close this text box to see the slide...

dobney.com

If there was a general election held tomorrow, which party would you vote for? Or would you not vote at all (Filter: females)
 

EU referendum preference

Base: 1525

All

Remain a member of the European Union
36
Leave the European Union
38
Undecided
22
Would not vote
4
Base: 505

South

Remain a member of the European Union
32
Leave the European Union
39
Undecided
24
Would not vote
5
Base: 152

London

Remain a member of the European Union
51
Leave the European Union
26
Undecided
23
Would not vote
1
Base: 270

Midlands

Remain a member of the European Union
33
Leave the European Union
41
Undecided
23
Would not vote
3
Base: 386

North

Remain a member of the European Union
34
Leave the European Union
40
Undecided
20
Would not vote
5
Base: 134

Scotland

Remain a member of the European Union
49
Leave the European Union
31
Undecided
19
Would not vote
2
Base: 78

Wales

Remain a member of the European Union
35
Leave the European Union
40
Undecided
19
Would not vote
6
X The crucial referendum question. When weighted, the data was pointing to a remain vote. The raw unweighted data was for leave.

Close this text box to access the chart...

dobney.com

As you may know the United Kingdom will have a referendum on its membership of the European Union on the 23rd of June this year. How will you vote on the question - Should the United Kingdom remain a member of the European Union or leave the European Union? (Filter: females)
 

State of the world

Base: 1525

All

Getting worse
64
Getting better
5
Neither
26
Don't know
5
Base: 505

South

Getting worse
64
Getting better
5
Neither
26
Don't know
6
Base: 152

London

Getting worse
52
Getting better
8
Neither
34
Don't know
6
Base: 270

Midlands

Getting worse
66
Getting better
6
Neither
23
Don't know
6
Base: 386

North

Getting worse
65
Getting better
5
Neither
26
Don't know
4
Base: 134

Scotland

Getting worse
71
Getting better
4
Neither
20
Don't know
4
Base: 78

Wales

Getting worse
64
Getting better
9
Neither
23
Don't know
4

dobney.com

All things considered, do you think the world is getting better or worse, or neither? (Filter: females)
 

Immigration view

Base: 1525

All

Mostly good
11
Mostly bad
43
Both good and bad
40
Don't know
6
Base: 505

South

Mostly good
10
Mostly bad
43
Both good and bad
41
Don't know
6
Base: 152

London

Mostly good
24
Mostly bad
27
Both good and bad
42
Don't know
7
Base: 270

Midlands

Mostly good
11
Mostly bad
47
Both good and bad
36
Don't know
6
Base: 386

North

Mostly good
6
Mostly bad
45
Both good and bad
41
Don't know
8
Base: 134

Scotland

Mostly good
14
Mostly bad
42
Both good and bad
41
Don't know
3
Base: 78

Wales

Mostly good
5
Mostly bad
45
Both good and bad
42
Don't know
8

dobney.com

And on balance, do you think the level of immigration into Britain over the last ten years has been mostly good or mostly bad for the country, or both good and bad? (Filter: females)
 

Response rates

The charts we will want to draw attention to for next time are the two following on likelihood to answer the phone and likelihood to take a call on a mobile.

In the next demonstration we'll run some calculations to show how propensity to answer can cause problems with response rates and representativeness within samples.

You might want to think about how this might affect telephone research for next time...

(Filter: females)

dobney.com

dobney.com

 

Phone answering - landline

Base: 1131

All

Always answer
24
Mostly answer
32
Mostly do not answer
19
Never answer
9
Do not have a landline telephone
15
Don't know
1
Base: 390

South

Always answer
25
Mostly answer
34
Mostly do not answer
22
Never answer
6
Do not have a landline telephone
13
Don't know
1
Base: 121

London

Always answer
17
Mostly answer
31
Mostly do not answer
17
Never answer
9
Do not have a landline telephone
26
Don't know
1
Base: 181

Midlands

Always answer
27
Mostly answer
30
Mostly do not answer
14
Never answer
10
Do not have a landline telephone
18
Don't know
1
Base: 292

North

Always answer
26
Mostly answer
29
Mostly do not answer
20
Never answer
10
Do not have a landline telephone
13
Don't know
1
Base: 94

Scotland

Always answer
20
Mostly answer
30
Mostly do not answer
21
Never answer
15
Do not have a landline telephone
13
Don't know
1
Base: 53

Wales

Always answer
21
Mostly answer
42
Mostly do not answer
13
Never answer
11
Do not have a landline telephone
13
Don't know
0

dobney.com

How often, if ever, do you answer phone calls to your landline telephone? Do you always answer, mostly answer, mostly do not answer, or never answer. (Filter: Contacted by mobile or online) (Filter: females)
 

Phone answering - mobile

Base: 1525

All

Always answer
8
Mostly answer
20
Mostly do not answer
29
Never answer
30
Do not have a mobile phone
4
Don't know
1
Base: 505

South

Always answer
7
Mostly answer
22
Mostly do not answer
30
Never answer
29
Do not have a mobile phone
5
Don't know
1
Base: 152

London

Always answer
6
Mostly answer
29
Mostly do not answer
29
Never answer
28
Do not have a mobile phone
3
Don't know
1
Base: 270

Midlands

Always answer
12
Mostly answer
17
Mostly do not answer
30
Never answer
31
Do not have a mobile phone
6
Don't know
1
Base: 386

North

Always answer
8
Mostly answer
18
Mostly do not answer
27
Never answer
27
Do not have a mobile phone
4
Don't know
2
Base: 134

Scotland

Always answer
10
Mostly answer
19
Mostly do not answer
33
Never answer
33
Do not have a mobile phone
0
Don't know
0
Base: 78

Wales

Always answer
5
Mostly answer
18
Mostly do not answer
27
Never answer
41
Do not have a mobile phone
8
Don't know
0

dobney.com

And how often, if ever, do you answer phone calls from telephone numbers you do not recognise on your mobile phone? Do you always answer, mostly answer, mostly do not answer, or never answer. (Filter: females)
 

Age and Gender

Base: 1525

All

Male 18-24
0
Male 25-49
0
Male 50-64
0
Male 65+
0
Female 18-24
10
Female 25-49
36
Female 50-64
24
Female 65+
29
Age and Gender (Filter: females)
 

Summary

This concludes our brief warm-up demonstration of live charting.

In the main Rethinking Questionnaire demonstration that will follow, we will move this much further to look at how Questionnaire techniques can be blended with charts and results and automated summaries to improve the reporting and communication process.

You can follow our full set of demonstrations on Rethinking Questionnaires at our website, or via LinkedIn .

(Filter: females)

dobney.com

dobney.com