In December 2016, the Daily Mail ran a story under the headline “Trust in politicians near rock bottom, survey reveals”. The story reported that politicians were now trusted to tell the truth by only 15% of the British public.

This story was based on a survey called the ‘Veracity Index’ conducted by the survey company Ipsos MORI. A full breakdown of the results is available from the Ipsos MORI website :

Look for the PDF containing the data tables.

Use the data in the PDF to answer the following questions:

1. Is the 15% figure reported by the Daily Mail correct?

2. Does the denominator for this calculation include people who say they ‘don’t know’ whether they trust politicians or not? Do you think it should or shouldn’t?

3. What percentage of young (15–24) vs. older (65 ) people trust politicians to tell the truth?

4. What percentage of people who trust politicians are young (15–24) vs. older (65 )?

5. Which of the previous answers is most useful in determining how trust in politicians differs by age?

The US General Social Survey is a large-scale survey that takes place every two years. The data from the survey is available for anyone to look at using the GSS Online Data Explorer.

You will have to register to use the site. However, registration is free. You can find out more information about registering (including terms and conditions) at the site itself.

Once you have registered and signed in, you can use the ‘Search GSS Variables’ function to look through all the survey questions that have been asked through the years.

For the purposes of this exercise we are going to look at data from the 2016 survey. First we will focus on the following question about homosexuality:

You can find this variable by entering ‘homosexual sex relations’ into the search field.

Use information provided to determine what percentage of respondents think that homosexuality is wrong.

Note: For the purposes of calculating the denominator, ‘not applicable’ means people who were not asked this particular question in the survey.

To answer this question, you will need to decide which response categories to group together.

Which categories did you include and why?

Next we are going to use the GSS ‘analyze data’ function to determine whether men or women are more likely to think that homosexuality is wrong.

To use this feature you will need to add the ‘homosexuality’ variable to your ‘cart’. This is like an online shopping cart that includes all the variables you want to work with. You will also have to add the variable indicating the respondents’ gender to the cart. You can find this variable by entering ‘sex’ into the search field.

Once both variables are in your cart, click the orange ‘actions’ button, then click ‘analyze data’. Now select ‘cross tabulation’ (you should remember from the chapter what this means), and click ‘next’.

On the next screen follow the instructions to indicate which variables will form the rows and columns of your table. Once you’ve done this, click the ‘create analysis’ button to produce the table.

Use the data in the resulting table to determine whether men or women are more likely to think that homosexuality is wrong.

The following is a hypothetical claim from a news story:

“In society, women bear a much greater burden of anxiety than men: figures from the UK Office for National Statistics show that the risk of anxiety is 30% higher for women than it is for men.”

This claim is based on data from 2,048 people participating in a nationally representative survey. Participants in the survey were asked the following question about anxiety:

Responses are given on a scale from 0 (lowest anxiety) to 10 (highest anxiety). The 30% figure is based on classifying people who give a reply of 8 or above as ‘experiencing anxiety’.

The raw data on which the calculation is based is available to download as an Excel spreadsheet here.

Use the data to answer the following questions (instructions for how to carry out the relevant operations in Excel are given here):

1. Based on the ‘8 or above’ cut-off. What is the risk of experiencing anxiety for men?

2. What is the risk of experiencing anxiety for women?

3. Is the “30% higher” figure accurate?

4. Do you agree that a response of 8 or above is the best indicator of experiencing anxiety? If not, what cut-off would you suggest?

What difference does this make to the relative risk figure for women?