According to the UK's Office of Communications (2015), young people between the ages of 16 and 24 spend more than 27 hours a week on the Internet. The average adult spends more than 20 hours online a week, which includes time spent on the Internet at work.

More and more, technology is an essential part of classrooms around the globe - let it be one-to-one laptop programs, use of mobile devices, Skype conversations between schools around the world, or digital textbooks (like the one that you are reading).  Programs like the Khan Academy have made personalized learning possible - that is, students are able to learn at their own pace by interacting with an online app.

The question psychologists want to answer is - is this good for us?

Although computers and the Internet have been around for some time, it is only now that we are beginning to see what we believe may be the effects of living so much of our lives in the digital world.  In the past few years, we have seen the following headlines:

• If you are going to take notes, do it by hand!
• New study links Facebook to depression
• In our digital world, are young people losing the ability to read emotions?
• Internet use may harm teen health: Scientists find that heavy Internet use puts teens at risk for high blood pressure

The study by Mueller and Oppenheimer used an experimental design and a volunteer sample.  But as you can see, it was not very natural.  The lectures were disconnected from the interests of the students, the participants were questioned a week later without being allowed to access their notes, and the trivia-style questions that were asked all made for a rather artificial experiment.  Field studies are also being used in psychology to study the effect of technology on our cognitive processes.. However, attempts to avoid demand characteristics by carrying out field experiments have raised serious ethical concerns with regard to informed consent.

Interest in how we study the impact of the digital world on the way we think, the decisions we make, and our emotional state intensified following the publication of a study often referred to as a ‘secret experiment’ conducted by Facebook.

Many people were angry that the study (Kramer et al, (2014)) had manipulated the information they received in their Facebook feed in order to test how this slight manipulation might affect them emotionally. Despite widespread and high-profile criticism, Kramer responded with a passionate defense of the study.

This research raises important questions about the way in which research should be conducted as well as the extent to which corporations should be bound by the same rules as academics. The study was conducted as part of an internal investigation within Facebook, thereby falling beyond the scope of the ethics committee at the relevant university. Facebook confirmed that the manipulation fell within their own Data Use Policy. It is worth considering the fact that everything that appears in your Facebook feed does so.

The manipulation of the news feed was done remotely and without any direct involvement of any researchers or Facebook staff. All participants were still able to see all posts if they took the time to view their friend’s ‘wall’. Kramer has argued that the research was important given the scale and intensity of social media use, especially Facebook.  However, as none of the participants were aware of any manipulation, there is little reason to suppose they would do so. Kramer argues that the benefit of such research outweighs the costs of failing to provide any form of consent.

Survey research

Much research into the way we are affected by the digital world has been conducted using survey data, often collected online. This method has several benefits, most notably the absence of any significant ethical concerns as participants are often fully informed or are at least provided with a full debrief and the right to withdraw their data if they are unhappy when the study is completed. However, research conducted with surveys consistently raises concerns in terms of both internal and external validity.

One limitation is the social desirability effect. Even when survey data is collected anonymously it remains true that many participants will be reluctant to provide any information about themselves which could be negatively judged. For example, participants may feel that is not socially desirable to spend too much time online and may, therefore, underestimate the average time they spend in online activity. It is also likely that people are unwilling to honestly describe what they are doing when they are online; the use of pornography and online gambling is certainly likely to be under-represented in any survey seeking to measure Internet use!

Another limitation is sampling bias. Surveys are often sent out or made available to very large numbers of people on the assumption that only some of them will choose to respond. This is very sensible from a practical point of view, (especially if surveys can be distributed without cost by electronic means) but it can reduce the population validity of the research. We might consider why some people would take part in a survey about social media while other people choose not to do so. For example, is it likely that those taking part have a particular interest in social media (or a particular aversion to it)? In either case, this may make it difficult to generalize results to a wider population.

Checking for understanding