The Rosie Project (Don Tillman #1)(8)

‘Relax, Don, plenty of time,’ he said. ‘We’re going to have a civilised dinner, and then we’ll do the questionnaire. If you’re going to be dating, you need dinner practice.’

He was, of course, right. Claudia is an excellent cook and Gene has a vast collection of wines, organised by region, vintage and producer. We went to his ‘cellar’, which is not actually below ground, where he showed me his recent purchases and we selected a second bottle. We ate with Carl and Eugenie, and I was able to avoid small talk by playing a memory game with Eugenie. She noticed my folder marked ‘Wife Project’, which I put on the table as soon as I finished dessert.

‘Are you getting married, Don?’ she asked.

‘Correct.’

‘Who to?’

I was about to explain, but Claudia sent Eugenie and Carl to their rooms – a good decision, as they did not have the expertise to contribute.

I handed questionnaires to Claudia and Gene. Gene poured port for all of us. I explained that I had followed best practice in questionnaire design, including multiple-choice questions, Likert scales, cross-validation, dummy questions and surrogates. Claudia asked for an example of the last of these.

‘Question 35: Do you eat kidneys? Correct answer is (c) occasionally. Testing for food problems. If you ask directly about food preferences, they say “I eat anything” and then you discover they’re vegetarian.’

I am aware that there are many arguments in favour of vegetarianism. However, as I eat meat I considered it would be more convenient if my partner did so also. At this early stage, it seemed logical to specify the ideal solution and review the questionnaire later if necessary.

Claudia and Gene were reading.

Claudia said, ‘For an appointment, I’m guessing (b) a little early.’

This was patently incorrect, demonstrating that even Claudia, who was a good friend, would be unsuitable as a partner.

‘The correct answer is (c) on time,’ I said. ‘Habitual earliness is cumulatively a major waste of time.’

‘I’d allow a little early,’ said Claudia. ‘She might be trying hard. That’s not a bad thing.’

An interesting point. I made a note to consider it, but pointed out that (d) a little late and (e) very late were definitely unacceptable.

‘I think if a woman describes herself as a brilliant cook she’s a bit up herself,’ said Claudia. ‘Just ask her if she enjoys cooking. Mention that you do too.’

This was exactly the sort of input I was looking for – subtle nuances of language that I am not conscious of. It struck me that if the respondent was someone like me she would not notice the difference, but it was unreasonable to require that my potential partner share my lack of subtlety.

‘No jewellery, no make-up?’ said Claudia, correctly predicting the answers to two questions that had been prompted by my recent interaction with the Dean.

‘Jewellery isn’t always about appearance,’ she said. ‘If you have to have a question, drop the jewellery one and keep the make-up. But just ask if she wears it daily.’

‘Height, weight and body mass index.’ Gene was skimming ahead. ‘Can’t you do the calculation yourself?’

‘That’s the purpose of the question,’ I said. ‘Checking they can do basic arithmetic. I don’t want a partner who’s mathematically illiterate.’

‘I thought you might have wanted to get an idea of what they look like,’ said Gene.

‘There’s a question on fitness,’ I said.

‘I was thinking about sex,’ said Gene.

‘Just for a change,’ said Claudia, an odd statement as Gene talks constantly about sex. But he had made a good point.

‘I’ll add a question on HIV and herpes.’

‘Stop,’ said Claudia. ‘You’re being way too picky.’

I began to explain that an incurable sexually transmitted disease was a severe negative but Claudia interrupted.

‘About everything.’

It was an understandable response. But my strategy was to minimise the chance of making a type-one error – wasting time on an unsuitable choice. Inevitably, that increased the risk of a type-two error – rejecting a suitable person. But this was an acceptable risk as I was dealing with a very large population.

Gene’s turn: ‘Non-smoking, fair enough. But what’s the right answer on drinking?’

‘Zero.’

‘Hang on. You drink.’ He pointed to my port glass, which he had topped up a few moments earlier. ‘You drink quite a bit.’

I explained that I was expecting some improvement for myself from the project.

We continued in this manner and I received some excellent feedback. I did feel that the questionnaire was now less discriminating, but was still confident it would eliminate most if not all of the women who had given me problems in the past. Apricot Ice-cream Woman would have failed at least five questions.

My plan was to advertise on traditional dating sites, but to provide a link to the questionnaire in addition to posting the usual insufficiently discriminating information about height, profession and whether I enjoyed long walks on the beach.

Gene and Claudia suggested that I also undertake some face-to-face dating to practise my social skills. I could see the value of validating the questionnaires in the field, so, while I waited for online responses to arrive, I printed some questionnaires and returned to the dating process that I thought I had abandoned forever.