Quantitative methods include formalized principles that form the basis for a stringent research process that proceeds from formulation of research questions, research design and the selection and analysis of data to interpretations and conclusions. The data will be linked to specific variables, and standardized methods are applied for data collection (for example in expert assessments, observations, interviews or formal testing). The variables can thus be expressed in numerical form, and the data material can be described in the form of tables, graphs or statistical measurements such as averages, variances and correlations, and analysed with the aid of e.g. analysis of variance, factor analysis or regression analysis (Befring 2015:131–160).
Even though quantitative methods are characterised by stringent requirements for structure, the methods also provide room for flexibility and pragmatic adaptation. Combined designs using "mixed methods" that include qualitative as well as quantitative data are therefore increasingly used (Cresswell 2014:215–240). This may imply that a qualitative study is included as a pilot study, or that qualitative data are used as the basis for a subsequent quantitative analysis. In addition, the data collection may be undertaken qualitatively (e.g. in the form of essay assignments) while the analysis is undertaken quantitatively. In this case, the information in the essays will be quantified and prepared for statistical processing.
Moreover, qualitative methods can be used to elaborate on findings from a survey. While surveys are able to provide a general overview of the matter at hand, qualitative data have the potential to provide more detailed insight into the opinions and experiences of the informants.
It is an essential requirement for researchers to have fundamental ethical attitudes that permit them to conduct all stages of the research process in an honest and credible manner.
Traditionally, the search for truth has been the fundamental objective and legitimisation of science. According to the "Galilean imperative" (from Galileo Galilei's formulation from around 1600), research shall investigate everything, reveal all mysteries, penetrate the unknown and provide objective explanations of all phenomena. Research shall not be governed by prevailing opinion, but search for true knowledge with no concern for other interests.
Another key value, which is closely related to the first, emphasises the need for academic freedom. The value of freedom, which has enjoyed a strong position, has been based on trust in the impartiality of research in contentious issues of a political, moral and religious nature. One key premise was that research could only be valuable when the researchers and their institutions remained neutral with regard to values and were unfettered in their social position. This perspective has changed in later times, since it is taken for granted that various values, issues and factors, such as formal arrangements and financial priorities, will influence the research that is undertaken. The question is therefore what kinds of values and socially normative factors in reality serve to govern research, and what values help justify and legitimise it.
One academic value concerns reliability and credibility, focusing on the quality of the research methods that are applied and the researchers' ethical standards.
A value that is of more recent origin states that research must be accessible for scrutiny and published, which is related to the need for control and possible unintended consequences.
Researchers are never exempted from exercising sound ethical judgement, they need to be methodologically qualified for conducting research with validity and quality, and they must do their utmost to ensure that objective concerns take precedence over preconceived notions and assessments.
Being honest means being fair and credible – being trustworthy. Dishonesty may sometimes manifest itself in plagiarism, fabrication or forgery of data, or in taking the credit for someone else's work. Plagiarism can include duplicating, copying or many other subtle forms of using the theories, interpretations, designs or results of others without reference to the sources.
Intentional falsifications give rise to fundamental problems of research ethics. These will include deliberate and planned acts to arrive at specific conclusions. As in the example above, there may be grounds to characterise this as cheating and fraud. In other words, this is scientific dishonesty in practice.
Misleading results may thus emerge when certain conclusions are of greater interest to the researcher than the scientific evidence that follows from an objective scientific investigation. This is the background for the questions raised about commercially based research and certain forms of commissioned research. When a piece of research work is used to corroborate certain arguments put forward by a commissioning agency, there is a risk of crossing the line that separates research from legitimising and manipulatory activities.
Unintended errors may be a direct consequence of inadequate competence in research methodology. In order to maintain the high-quality work that research implies, it is thus a key precondition that the researchers are competent to use relevant and acceptable research methods.
Protection of integrity
Clear initiatives have been taken to prevent unworthy conditions in the form of scientific misconduct, academic dishonesty etc., and a number of measures have been enacted to strengthen research ethics as a pillar of scientific practice. In psychology and educational science, the American Psychological Association (APA) has played a key role through its Ethics Code, which since the early 1980s has been published in increasingly complete versions. These ethical norms include issues such as requirements for honesty, requirements for informed consent, anonymisation and storage of data, the right of access to data for participants and duty of confidentiality for all those who undertake research.
Protection of the integrity of research participants and informants is a particularly important ethical norm in research, including in special-needs education studies. This norm focuses on protection against various forms of risk involved in participation in research and the protection of the identity of participants, including concerns for preventing stigmatisation of particular populations or groups. For example, such a risk will be present if it transpires that a particular ethnic group is overrepresented among families and children who have been reported to the child protection services. In such cases, it will be essential to clarify the background, while at the same time emphasising that the child protection services are an integral element of a welfare state. In any event, publication of such forms of research involves an ethical challenge.
Protection of anonymity and thus privacy is a key issue in protecting integrity. However, the practical application of this norm may come into conflict with a wish to use a specific design. This may involve the need for methodologically valid replications and testing of research findings, and there may be a need for linking data from various sources. The latter may include, for example, questionnaire data from students that the researcher wishes to link to the teachers' assessments of the same students.
It has also been shown that the use of electronic methods as well as telephone interviews often gives rise to scepticism regarding real preservation of anonymity. As a result, certain informants will refuse to participate (Coen et al. 2011:94–95).
In addition, it will generally be more difficult to protect the anonymity of informants in qualitative studies than when collective, quantitative methods are used (such as various forms of questionnaire-based methods). Irrespective of whether this involves observations or interviews, the individuals will stand out in a direct and visible manner. In personal interviews, the informant will recount previous experiences in his or her own words, which may be recognisable to others. When designs for collective questioning are used, however, individual characteristics will be less identifiable, and individuals who have completed a questionnaire will not be directly visible.
Control and ethical competence
It is often assumed that the standardized methods and formalised requirements for quantitative research help ensure academic and ethical credibility. Even though there will be fewer subjective elements of error here than in qualitative research, there will be ample opportunities for dishonesty to render the research misleading even when quantitative methods are applied. At worst, researchers may manipulate the data, use fictitious and fabricated data or discard any unwanted results. When research is undertaken in solitude and the researcher is alone in having insight into what is being done, the road to dishonesty lies open.
Numerous examples of research fraud can be found, in both an historical and contemporary context. These examples include academics who were held to be researchers of high stature, but in reality were untrustworthy. In educational psychology, Cyril Burt (1883–1971) represents an especially grave case. By fabricating and manipulating research data in his studies of twins, he could "confirm" his theory of heritability of intelligence. International academic literature contains innumerable descriptions of "The Cyril Burt Affair". The essence of Burt's reprehensible acts consisted in the fact that many of the twins in his material did not exist, nor did many of the researchers with whom Burt had "collaborated".
However, this scandal draws attention to the fact that the research community has traditionally been closed and disinclined towards openness, including with regard to potential risk factors, and thus underscores the need for a clear focus on values as well as for increased transparency and monitoring.
In light of what we know about dishonesty, two self-evident countermeasures are commonly identified: first, to strengthen the monitoring, and second, to raise the ethical standard. The question is what we can do to develop these areas further. There can hardly be any doubt that the monitoring is often of a superficial nature. Those who have funded research are often content to receive some research reports, and unannounced observations of the research process are rarely undertaken. In most cases, monitoring amounts to a critical reading of a report or an article. Ideally, however, replication by means of an empirical investigation of the key conclusions ought to be a standard requirement for publication. However, monitoring may be of limited value and may also entail unintended negative consequences, since the desired creativity presupposes freedom as well as trust.
Facilitating an expanded ethical competence in researchers and research communities is therefore a key concern. This includes ethical awareness and follow-up on the part of agencies that provide research funding. This is also a matter of ethics in a wider context, in the need to focus on power relationships and processes that maintain hegemony, whereby particularly subjects in the fields of care and learning are systematically discriminated against when resources are allocated. This is also a matter of those who repeatedly have resources allocated to them, the self-recruitment for such assignments and at worst, the "guarantors" of favourable conclusions. This topic has attracted little attention in this country so far, and it is also beyond the limits of this treatise.
Above, we have focused on how research competence includes ethical, professional and methodological credibility. For example, when educational researchers possess limited insight into learning and the learning processes in children, they will have very limited qualifications for undertaking adequate empirical research. Thereby, the ground is prepared for irrelevant and misleading conclusions. This is a hidden ethical problem in research.
This may also apply to the implementation of quantitative studies, often involving large amounts of data, which require methodological competence to conduct analyses in a stringent and academically acceptable manner.
A largely overlooked ethical problem is associated with the issue of choice of design in research that includes, for example, children in a difficult life situation, disabled people or others who are in need. For example, one may question the ethical defensibility of using an experimental design with control groups. Those who have a hard life are often burdened by many crushed hopes, and participation in a control group may thus be problematical in terms of ethics. Various single-group designs (time-series designs) may be preferable, not least in the context of special-needs education (Befring 2015:33).
This is a matter of being ethically considerate and maintaining good research practice, which can uphold research quality as well as concern for those people whom this research is intended to serve.
Another problem of research ethics is associated with publication without sufficient quality assurance and without replication. One especially unworthy and problematic aspect of this are media reports that trumpet dramatic conclusions, when no research report is available. Such media reports often cause a stir, and not infrequently they spark a debate that has no research base and is often misinformed. This practice serves to undermine the trust in research, and can thus be detrimental to research that rests on a solid academic and ethical foundation.
This article has been translated from Norwegian by Erik Hansen, Akasie språktjenester AS.