The first thing that may spring to mind when the subject of research ethics comes up is the need to look out for participants’ welfare – we cannot subject
them to just any conditions for the sake of answering our research question and this is discussed below. First, the issue of the relationship between science and society is considered (Diener & Crandall, 1978) due to its relevance to research on the TFD.
Science and society
The issue of the relationship between science and society concerns the extent to which society should influence which research topics are pursued (Christensen, 2000). Competition for research funding can be intense, and the money tends to go the projects that are ‘in vogue’. For example, our society is currently con- cerned about poverty in third world countries, the economic climate of our own country, obesity, and immigration (amongst other things). Correspondingly, the Economic and Social Research Council (ESRC), one of the primary research funding bodies in the UK, states in its priorities for the 2009-2014 period ‘redu- cing poverty among the poorest countries in the world’ (p. 6), ‘understanding of individual and household responses to the rapidly changing economic climate’ (p. 6), ‘the reduction of obesity’ (p. 8), and understanding ‘the dynamics of migration into and from the UK’ (p. 21, ESRC, 2009).
What this subsequently means is that researchers may alter their research focus for the sake of obtaining funding. While it is not necessarily wrong for researchers to investigate what is important to society, especially when research funding is coming from taxpayers’ money, it does compromise the objectivity of science. Traditionally, science is supposed to be the uncovering and explain- ing of the nature of our world by objective scientists but this may well be compromised when researchers have to compete for funding by tailoring their research questions to social or political motives. As Christensen (2000) points out, researchers’ interests may also be determined by personal experiences. For example, a dyslexic researcher may want to investigate dyslexia in school chil- dren, and again, this may compromise the objectivity of the research.
The issue of society influencing science is relevant to the research reported in this thesis. As discussed in Section 2.7 on the current status of the TFD, there have been many recent arguments for mathematics to be prioritised in the National Curriculum based on the assumption that studying mathematics improves thinking skills. Investigating the accuracy of that assumption is a part of the motivation for conducting this research. It is important, therefore, that the research is conducted as objectively as possible – not biasing the results towards either supporting or refuting the TFD. One way in which this issue is dealt with is by withholding the purpose of the study from the participants (discussed below). Another benefit to objectivity is that I am approaching the
question with a background in psychology, not mathematics, and so do not feel the need to support mathematics by finding the TFD to be true, whereas others with a background in mathematics might.
Welfare of participants
Participants in research studies should always be asked for informed consent, given the right to withdraw, and debriefed after participation. Wherever pos- sible, research should be conducted in a way in which participants do not suffer any negative effects, such as failure, stress or embarrassment, and they should not be deceived (Ethics Committe of the British Psychological Society, 2009). Two of these issues which have particular importance for the research presented in this thesis are informed consent and deception.
Informed consent means that participants should be told all relevant informa- tion about the nature and purpose of the research and that they should only subsequently take part having given their voluntary consent. When the par- ticipants are vulnerable adults or children special considerations are required for informed consent. The participants in all studies reported in this thesis, including in the AS level study, were aged 16 or over. According to the BPS (The British Psychological Society, 2010), only those under the age of 16 are considered to be children. Nevertheless, parental consent was obtained for the participants aged under 18 in Chapter 5. The participants in all of the stud- ies reported here were given information about the purpose and nature of the studies they were asked to take part in (except for some minor withholding of information in the case of the AS level study, see below), and all gave consent before their participation began.
Deception and withholding information
In the longitudinal studies reported in this thesis, it was necessary to withhold some of the details of the purpose of the study from participants. If they were aware that I was testing the hypothesis that mathematics students would im- prove in reasoning to a greater extent than the English/psychology students, the results may have been invalidated by stereotype threat.
Stereotype threat occurs in test situations when a person could potentially reinforce a negative stereotype about a social group they belong to. For example, when men and women were given the same mathematics test and told either that it did or did not tend to show a gender difference in scores, Spencer, Steele and Quinn (1999) found that gender differences in performance followed accordingly:
for the group of participants who were told the test did not produce a gender difference, no gender difference was found, but for the group that were told that it did produce a gender difference, women scored significantly lower than men. In another study, Asian-American women were found to perform better on a mathematics test when their ethnic identity was primed, and worse when their gender was primed, than a control group who had neither identity primed (Shih, Pittinsky & Ambady, 1999). White men’s mathematics performance was also found to suffer when they were reminded that Asian people tend to outperform white people on mathematics tests, compared to a control group that was not reminded of the stereotype (Aronson, Lustina, Good & Keough, 1999). These examples demonstrate what a powerful effect stereotype threat is – a single statement from researchers can create a significant difference in the performance of participant groups.
In the longitudinal studies reported here, stereotype threat could be a prob- lem if participants were told that I was testing the claim that mathemat- ics students would improve in reasoning to a greater extent than the Eng- lish/psychology students – it could be enough to create the difference between groups that I was testing for. Instead, participants were told that the research was looking at improvement in reasoning over A-levels without emphasising any comparisons between subjects. The participants were not deceived, therefore, rather information was withheld. In this case, the cost to participants was minimal, and the findings would simply not be valid if the information were not withheld. In line with the BPS code of conduct (Ethics Committe of the British Psychological Society, 2009), it was deemed that this withholding of information was appropriate, and the study was approved by Loughborough University’s Ethics Approvals (Human Participants) Sub-Committee.