It seems perfect: no queues, screaming children, students buying 10 Marlboro Lights, people bickering about the price, stroppy assistants, wonky trolleys, six-packs with a third off, old strawberries with new discounts. Shopping on the internet should be easy and stress-free, yet, says a new study, people are still reluctant to embrace e-commerce.

Dr Peter Lunt, senior lecturer in the psychology department, University College London, has spent the past two years analysing people's responses to internet shopping and found that most shoppers buy books, CDs, travel tickets and software over the internet. He surveyed almost 900 people, talking to and surfing the net with groups and individuals in their own home.

Of those surveyed, only 14%had tried shopping online and a mere 5% were regular (monthly) users. Three-quarters had bought only four kinds of products. In general, the people least keen on internet shopping tended to be thrifty or brand loyal shoppers.

According to Lunt, the main reasons cited for being cautious about e-shopping were the costs of computer equipment, going online and delivery of goods. People were also uncertain about how to return products.

There is also a fundamental mismatch between the way people shop normally and online services: impulse buying or the ability to examine products are missing; for many, the experience of shopping, for clothes, for instance, is as important as the products themselves.

"People acknowledge the convenience of e-commerce in principle, especially for mundane grocery shopping," says Lunt. "But it became clear that the supermarket shop is a complex activity that includes idiosyncratic and luxury items bought on impulse or with a specific occasion, person or meal in mind. The difficulties of reproducing this experience using a list-based computer program were evident to our participants."

Most shoppers found it hard to navigate e-commerce sites as goods are laid out unintuitively and many were disappointed at how dull the web sites were, particularly if they were used to computer or video games.

Lunt believes there are three groups of people who do not e-shop. The first has little knowledge of the internet but is potentially interested. "They're a prime target for limited e-services delivered by digital TV," he adds. The other group tends to be older and less educated and, Lunt believes, may be left behind. Members of the third group are relatively wealthy and computer literate but have other reasons for not shopping on the web. "The issues for these people are the broader social issues of privacy and the possible effects on the way we live and shop. They could be the new Luddites - except that they are happy to use the internet for other reasons."

Lunt adds that this group is unlikely to start e-shopping unless they are sure that their privacy will be safeguarded.

Most of the people he talked to considered online shopping analogous to "real" shopping. "They are thinking more of the integration of e-commerce services into their current consumption practices rather than taking the opportunity to rethink the way they shop and organise the allocation of responsibilities within the household," says Lunt.

The location of the computer was another factor, as many people have theirs either in the sitting room as part of an entertainment set-up, or else in the study for work.

Lunt sees this as a drawback because as he puts it, "PCs are not integrated into the 'provisioning' spaces and flows of the household".

Young people and men were most positive about internet shopping. About 58% of e-shoppers are men. Lunt suggests that men shop more than women simply because what they want is gendered - books and CDs rather than cosmetics or clothes - but that in any case men still aren't buying much at all.

"The survey indicates that those who are most positive about e-commerce are focusing on the convenience and the price advantage in particular products and not on the experiential aspects of shopping. These are relatively well to do people who are technologically savvy and don't get much pleasure from shopping," he says.

Given that these people are a minority, will the situation change enough to encourage the rest of the population to shop online? The European Commission has already been involved in producing directives aimed at protecting the identities of online shoppers, but the trouble will be in regulating the internet.

As usual it will be down to the manufacturers themselves to behave reasonably. How ever, because manufacturers will need to make more profits than usual from internet sales, they will need - and want - to collect and analyse the vast amounts of data that will be generated by internet shoppers.

"One likely solution is that firms will develop methods that allow the collection of data linked to consumer profiles (ways of categorising consumers using a variety of indicators) that do not require the identification of individuals," says Lunt. "This has the advantage of protecting the privacy of the individual but allowing the collection of detailed consumer information. The problem with this scenario is that this is still a radical increase in information asymmetry between consumers and retailers (they may not know it is me but they know a lot more about people 'like' me) and that there are a number of potential dangers in this practice. One is that marketing decisions will be made on the basis of the information but (since it is not identified with me) I can not challenge these decisions."

In the future, more of us may be encouraged to shop online as manufacturers overcome bad web design and learn to safeguard our privacy, but could internet shopping ever seriously replace the stress, and the thrill of the real thing?

"It depends how far you are looking ahead," says Lunt, "I think the interesting question is what the strategic responses of firms to the introduction of e-commerce will be. They have to find new ways of appealing to customers and offering services that will include having a mix of online and offline presence in the market . This raises interesting questions for existing firms: how will this affect their brand image?"