Subjectivity in Quantitative Nursing Research: Supporting the Postpositivist Views

Mohammad A. Al-Motlaq (1) Ysanne Chapman (2) (1) Dr. Mohammad A. Al-Motlaq, PhD, MBS, RN, BSN School of Nursing, Hashemite University, Jordan (2) Professor Ysanne Chapman RN, PhD, MSc (Hons), BEd (Nsg),GDE,DNE,DRM, MRCNA Dean, School of Nursing and Midwifery CQUniversity Australia, Mackay Campus Correspondence: Dr. Mohammad A. Al-Motlaq (Corresponding Author) PhD, MBS, RN, BSN School of Nursing, Hashemite University, Jordan Telephone: (05) 3903333, Fax: (05) 3826613 Email: MohammadA_M@hu.edu.jo

Abstract The credibility of research is important for its consumption. Reviewers and consumers of research place huge emphasis on the validity and reliability of the results which are mainly affected by the level of objectivity in the research process. Debates still exist between supporters of different paradigms as they disagree on what makes research credible, what makes it valid and to what extent, and which methodology is more appropriate. This paper presents a comparative view of credibility between quantitative and qualitative paradigms by exploring the issues of subjectivity and objectivity in its methodologies. In this discussion more weight is placed on the "subjective decisions" made by the so called "objective researchers". Key words: subjectivity, objectivity, quantitative and qualitative research, methodology, nursing

INTRODUCTION
The main purpose of research is to discover a reality that is believed to exist through exposition and discussion. Researchers have various definitions of reality dependent on their philosophy and experience. Metaphysically speaking, Positivists believe that reality is fixed and orderly and can be uncovered objectively, while Postpositivists believe it is imperfectly comprehensible and never perfect (1,2). Thus, in quantitative research the credibility of knowledge is maximized when the distance between the researcher and the subjects is minimized (3) placing the subject(s) within a highly objective state where no subjective interference can affect the findings. Postpositivists, however, believe findings are always subject to falsification by pre-existing knowledge or critical others, such as editors (1,2), and in a probabilistic sense (4). In terms of objectivism and relativism, quantitative inquiry supporters have long claimed it to be foundational and capable of describing reality as it really is (5, 6) though threats usually exist to falsify them. For them, what is so special about science is that it is derived from facts and not based on personal opinions (7). Qualitative researchers, however, assume that reality is subjective (highly affected by the researcher) and that multiple realities exist (4) as evident, for example, by the introduction of the Q-methodology (8). Although appearing often in the context challenges to methodology (9), the consistent debate between the two methodologies has surpassed the broader umbrellas of ontology and epistemology. It is inappropriate to use the term methodology in discussion of ontological claims (9) since the philosophical underpinnings of research methods are rarely explicated in nursing literature (10). Exact translation of the divergent validity of the two approaches is not appropriate (11) though translated standards of validity are legitimate (12). To distinguish between the two paradigms, authors have long argued the credibility of approaches based on their beliefs of what can be known and what methodology is more appropriate. While debates still exist, supporters of each paradigm appear to be no closer to a consensus (13). In addition to being the default research modality (14), supporters of quantitative methodologies base their arguments of credibility on the objectivity of the inquiry. In nursing research, however, reaching a highly objective quantitative study is impossible because of the many subjective decisions made by researchers. This article presents evidence from the literature on the subjective decisions quantitative researchers take in their "objective studies" and discuss their effect on the findings credibility.

Subjective decisions made through the scope of an "objective study"
For quantitative researchers, the research process starts with acknowledgment of a problem. This is in itself a subjective recognition of a dilemma and a biased acceptance of an issue. A well known example is the recognition of the climate change problem (Does it Exist?). The following sections present some of the subjective decisions made in quantitative nursing studies which are thought to affect their credibility, supported with examples from the literature.

Subjectivity in reviewing the Literature
Unlike qualitative researchers who avoid an in-depth literature search not to contaminate their inquiries by previous knowledge (3), quantitative researchers undertake a thorough literature search to establish the problem, largely to determine any gaps in the knowledge, and identify a suitable study design that assists analysis of their research findings. These purposes can be affected by individual decisions and are not fixed with one rule. Researchers decide what to read and what to exclude, and what databases to consult and what type of sources (primary or secondary) to rely upon. In addition, there is a strong reliance on one language, usually the language of the researcher; the Chinese medical experience is a standing example. Some Chinese drugs have found recognition in Western Medicine, (e.g., the treatment of Malaria). In one study, for example, an RCT proved Chinese herbal formulation improved symptoms for some patients with Irritable Bowel Syndrome (15).

Although some systematic reviews have been very exhaustive, no simple guide is available to accept and interpret findings. For the same question, multiple designs and inclusion criteria can be used. For example, a systematic review on stress management interventions for mental health nurses (16) which included papers that only met certain criteria (English language publication, between 1966 and 2000, professional groups concerned, primary research, and measuring specific outcomes) and excluded others (foreign language publication, papers with insufficient statistical data, and studies with subjects other than psychiatric nurses) based on a subjective judgement by the reviewers; as reviewed papers were reconsidered where there were differences between the reviewers judgment. One can hardly find a nursing study based on absolute evidence or systematic review/Meta analysis. For example, Meijers et al. reviewed the literature with the aim to examine relationships between contextual factors and research utilization in nursing and found no single study to be of high methodological quality for the purpose of their review (17). In addition to the experimental references, researchers might base their studies on opinion articles or anecdotal data, bearing in mind that evidence is not always evaluated, let alone the high percentage of errors in presenting relevant data (18). In turn these misinterpreted findings become accepted 'truths' which are perpetuated each time they are used in literature reviews, falsifying subsequent research.

Subjectivity in seeking ethical approval
Research with human subjects requires ethical approval. Amendments and changes could be required in some or all research elements if the ethics committee of the day suggests the research could predispose subjects to harm or might not protect their privacy. Ethics committees and their rules are dynamic; they have to adequately respond to the pressures of the day. Deciding the rules that make the code of ethics in research is subject to human decisions and the process, whilst carefully debated and proposed, is not a highly objective issue. This is evident by the many changes made to the rules and regulations of health research as well as the differences that exist between countries, particularly with studies involving vulnerable subjects. The way any scholar, philosopher or ethicist conceptualizes ethics and interprets the meaning of the (moral) language is not universal in nature because conceptions of an ideal moral judgement differ (19). Ethical regulations of research with humans affect the choice of design and might change the line track of experiments. Although ethical requirements for conducting medical research in developing countries have achieved considerable prominence in recent years (20), many studies are conducted in some countries that simply could not be undertaken in others (21).

Subjectivity in choosing a design
The research design incorporates most important methodological decisions researchers make (3). There are rules for excluding designs from top hierarchies, yet this by itself is a subjective matter and might vary between researchers. Designs lacking one or more conditions of a true experiment (manipulation, control, and randomization) might decrease the credibility of findings. Researchers have begun to employ mixed methods to solve practical research problems (22). Five purposes for mixed methods are identified, all of which hold within them, subjective significance: triangulation, complementarity, development, initiation, and expansion (23, 24), though others think mixed methods cannot be combined for cross-validation or triangulation, merely for complementary purposes (5). The choice of a design, rather than achieving congruency with the research question in some cases, is dependent on the experience of researchers, the resources available to them and their ability to conduct a study rigorously. Delphi technique, for example, has been utilised by many nurse researchers, and the rigour associated with the original format has been threatened. Keeney et al. critically examined the Delphi technique and found no one study used the `Delphi' in the same way; which, in their opinion, could be criticised as a threat to the uniformity of the method (25). Researchers' decisions about design are independent of those for data collection methods. Deciding which method is the 'best' for answering the research question may often fall outside the congruency argument and once again is dependent on previous knowledge of the researcher.

Subjectivity in choosing the sample
Quantitative researchers recruit samples that allow them to generalize their findings. However, a representative sample is the one whose features approximate those of the population and may not exactly match those of the population, let alone the non-probability sampling procedures used. The question here is; is 'approximate' good enough? In many cases this question is seldom asked and sample choice and their margins for error, glossed over. Although it is problematic, non-probability samples are used in most nursing studies because it is convenient and economic. Another subjective decision when choosing the sample is the power analysis, which mainly depends on estimates of the effect size. Results of a power analysis on 62 nursing articles (26) indicate that a large number of published nursing studies have insufficient power to detect real effects because of the small samples used.

Subjective decisions in collecting the data
While a number of data collection approaches entail more subjective judgment, other research problems require higher degrees of objectivity (3). Data collection tools are enormous in quantitative research, and the use of one tool over the other is subject to the researchers' evaluation of its validity and reliability. Although there are some rules, what is valid and reliable for one researcher might be not for another, as many researchers recognize the weakness in some tools yet still use them. Researchers might use self-reports to answer their questions, while others use observation to answer similar questions. For example, collecting data from people through surveys, interviews, or focus groups may provide useful information but this does not make the activity a research study (27). Another example is the use of a health diary in nursing research and the advantages and disadvantages associated with its utilization (28). As such, limitations in data collection tools include the vulnerability of data to researcher's biases (29).

Subjectivity in analysing the data
Two subjective decisions emerge in this regard; the subjectivity in the essence of statistics and the subjective decisions taken by the nurse researcher during the analysis. Although statisticians developed conditions and mathematical equations to rule data analyses, limitations still exist (30). Kenny et al. stressed the notion that data analysis should be a more thoughtful process as standard data analysis tools remain the same (30). Quantitative analysis of data relies basically on the theory of statistics, a theory that uses numbers to represent measured variables. As a matter of fact, the discrepancy between these numbers and the actual values of variables is known as measurement error. Researchers try to keep measurement error to a minimum by measuring the validity and reliability of their instruments which relies basically on subjective decisions. Cronbach's alpha, for example, a measure of reliability is considered acceptable if its value was 0.8 or more, though others subjectively say 0.7 or more (31).

Another example of the subjectivity in statistical analysis is regression where researchers try to explain how well a set of variables is able to explain a particular outcome (32). In regression analysis, researchers never reach a 100% explanation of a variance as there are undetectable variables that might predict the outcome variable as well. Likewise, Analysis of covariance might help remove the effect of an extraneous variable but certainly not all confounds.
Subjective judgments in dealing with data once back from subjects include cleaning, counting, and coding; these activities also involve decisions on missing values. Another subjective decision relates to testing hypotheses which requires gathering data about dependent and independent variables which are "thought" to have some kind of cause and effect relationships (31). Statistically testing a hypothesis has a number of conditions which are greatly affected by the researcher's opinion and not totally subject to objective rules. Researchers still engage in incorrect practices; using some tests when others should be used. Although some tests exist to guide decisions about normality of distributions, hence the use of appropriate statistical tests (31), researchers vary in their decision for considering a distribution normal or not. Researchers need to remember the basic assumptions of parametric analyses which include normality, linearity, multicollinearity, and homoscedasticity (32). In this regard, a review of quantitative methods used in health promotion research (33) found limited use of advanced statistical techniques that could help address important knowledge and practice issues. Statistical presentation often appears like a code which can deter nurses' full understanding of the evidence (34). Another study that reviewed the presentation and analysis of ordinal data in nursing research (35), found a large percentage of nursing studies do not present and analyse data properly, resulting in misleading information.

Subjectivity in interpreting the findings
The interpretation of findings is the major subjective decision a quantitative researcher can take. Prior to drawing conclusions, findings require objective and critical interpretation. However, many studies have been interpreted in a way that reflected the researcher's arrogance to an ideology and not the reality itself. Evidence for this is exemplified in the different reviews a researcher might gain from two independent reviewers; one positive and the other negative. Researchers need to remember that a lack of impact or effect is not sufficiently established by a failure to demonstrate statistical significance (36). Nevertheless, in some cases it has been interpreted as just that. The relevance of findings must be examined by considering alternative explanations such as concurrent influences, subjective measurement techniques, and statistical regression (37, 38).

Subjectivity in drawing conclusions and making recommendations
To be reasonable, recommendations and conclusions should be directly linked to the results. However, events sometimes happen that cannot be detected and subsequently influence the researcher's concluding statements and recommendations. For example, a review showed that despite the considerable work that had been done to establish the interpretability of quality-of-life measures much more work is left to be done on its acceptability (39). Guyatt et al. stress that the field remains controversial, and there are many alternative approaches, each with its own advocates (39). Accurate recommendations drawn from accurate conclusions are the only method that can ensure future practice is not repetition of previous failed attempts. If there are accurate evidence-based recommendations, future success is almost guaranteed.

DISCUSSION
Despite the increased recognition of qualitative research in nursing, debate on its objectivity and validity still exist. Consequently, articles on credibility and representativeness of qualitative research have been written (40-43) and the stigma attached to quantitative research is slowly being eroded by violation of the objectivity rules, the nature of nursing research, the availability of extensive research findings supporting the claims, and the increased support of specialist qualitative researchers in nursing. The aim of this article was to remind those who critique qualitative studies on their credibility that subjectivity also exists in quantitative research. Absolute objectivity in social and health research such as nursing is highly unlikely and the notion of having hierarchies of evidence by itself is a wiping out of the objective nature of quantitative research.

Drew discussed the gap between subjective experience of researchers and the inherent objectivism of science and research (44). She supports the views of Husserl about the danger arising from adopting only an objectivistic positivistic model of the world and ignoring the personal beliefs of researchers and how they experience themselves and their work (44). Drew, in the beginning of her discussion, claimed the absolute objectivity of quantitative research, a claim that was encountered by an opposite view later on when she presented the views of Husserl, asserting researchers return to their immediate experience and to the life world from which their enterprises arise (44). These views are supported in this article though for quantitative research. While the core purpose of research is looking for reality, this paper discussed the reality of nursing research when it explored what has happened, and what is happening and not the hypothesised ideal theory of objective quantitative nursing research.
This paper systematically presented examples of subjective decisions taken by quantitative researchers as an everyday practice that show threats to its core advantage over the qualitative paradigm of objectively uncovering the reality, a reality that is faultily apprehendable and never ideal. Despite these limitations, quantitative researchers still defend their approach with the claim that some empirical data are better than none. However, it is even now accepted by a significant number of both qualitative and quantitative researchers that qualitative research methods can be used to identify causal relationships and develop causal explanations (45). Based on the question, findings from qualitative research have a place in evidence based nursing practice, much the same as quantitative studies do (46).

The many articles critiqued within nursing research provide unambiguous substantiation on the limited rigour these studies had. The reviews on errors in nursing research, particularly quantitative, are clear evidence on the falsification of their conclusions. It is imperative to realize that scientific theories cannot be conclusively proved or disproved. If reasoning in drawing conclusions from factual basis is sound, which is debatable and subjective, the resulting knowledge can be considered objective (7). A literature review on utilization of nursing research (17) showed that nursing knowledge is not reflected in the practice of care, and researchers recommend more robust methods for better understanding of the impact of contextual factors on nurses' use of research. Monti and Tingen suggest multiparadigmism for the present and future development of the nursing science (47). Others (48) advocated that all graduate students should learn to utilize and to appreciate both methodologies as De Leeuw claimed that mixing modes has only advantages (49). Unlike some who thought it is not (6), the compatibility and cooperation between the two paradigms is sustainable.

CONCLUSION
The question of which paradigm should guide nursing science still stands. In the main, the objectivity of quantitative research leads the argument into its 'right' of being the bases for nursing research. However, quantitative studies are couched in subjectivity and the reality is still subject to researchers' judgment, and it is never perfect. The steps of quantitative research are subject to the researcher's own judgment of appropriates or feasibility which in turn depends on the allocated resources and experience. For consumers of research, whether they are ordinary people or experts, understanding of the origins of science gives them a way to judge the value of research. While each approach has its own advantages and disadvantages, its own strengths and limitations, there is a need for quantitative researchers to admit their subjectivity in the execution of their research and for qualitative researchers to recognise how quantitative research could add value to their research endeavours. Perhaps a mixed methods approach might be one feasible solution for tempering the debate.

REFERENCES
1- Guba E & Lincoln Y (1994) Competing Paradigms in Qualitative Research. In Handbook of qualitative research (Denzin N.K. & Lincoln Y.S. ed.), Thousand Oaks, CA, Sage Publications.
2. Wills J (2007) Foundations of Qualitative Research: Interpretive and critical approaches. Sage publications, Thousand Oaks.
3- Polit D & Beck C (2004) Nursing Research: Principles and Methods (8th ed.). USA: New York: Lippincott Williams and Wilkins.
4- Creswell J (2003) Research design: qualitative, quantitative, and mixed methods approaches (2nd ed.): Thousand Oaks: Sage Publications.
5. Sale J, Lohfeld L & Brazil K (2002) Revisiting the Quantitative- Qualitative debate: Implications for mixed method research. Quality and Quantity 36: 43-53.
6- Smith J & Heshusius L (1986) Closing Down the Conversation: The End of the Quantitative-Qualitative Debate Among Educational Inquirers. Educational Researcher 15(4): 4-12.
7- Chalmers A (2004) What is this thing called science? (3rd ed.): Biddles Ltd., London.
8- Akhtar-Danesh N, Baumann A & Cordingley L (2008) Q-Methodology in nursing research: A promising method for the study of subjectivity. Western Journal of Nursing Research 30: 759-773.
9- Rawnsley M (1998) Ontology, epistemology, and methodology: A clarification. Nursing Science Quarterly 11(1): 2-4.
10- Cody W (2002) The ontology-epistemology-methodology linkage: Still important after all these years. Nursing Science Quarterly 15(4): 274.
11- Bailey P (1996) Assuring Quality in Narrative Analysis. Western Journal of Nursing Research 18: 186-194.
12- Whittemore R, Chase S & Mandle C (2001) Validity in qualitative research. Qualitative Health Research 11(4): 522-537.
13- Rolfe G (2006) Validity, trustworthiness and rigour: Quality and the idea of qualitative research. Journal of Advanced Nursing 53(3): 304-310
14- Sandelowski M (2008) Justifying qualitative research. Research in Nursing and Health 31: 193-195.
15- Bensoussan A, Talley N, Hing M, Menzies R, Guo A & Ngu M. (1998) Treatment of Irritable Bowel Syndrome with Chinese herbal medicine: A randomized controlled trial. Journal of the American Medical Association 280: 1585-1589.
16- Edwards D & Burnard P (2003) A systematic review of stress and stress management interventions for mental health nurses. Journal of Advanced Nursing 42(2): 169-200
17- Meijers J, Janssen M, Cummings G, Wallin L, Estabrooks C & Halfens R (2006) Assessing the relationships between contextual factors and research utilization in nursing: systematic literature review. Journal of Advanced Nursing 55(5): 622-635.
18- Vaux D (2009) Ten rules for the presentation and interpretation of data in publications. A Lecture at Monash University, Clayton Campus.
19- Lützén K (1997) Nursing Ethics Into the Next Millennium: A context-sensitive approach for nursing ethics. Nursing Ethics 4(3): 218-226.
20- Benatar S (2002) Reflections and recommendations on research ethics in developing countries. Social Science and Medicine 54: 1131-1141.
21- Angell M (1997) The Ethics of Clinical Research in the Third World. The New England Journal of Medicine 337(12): 847-850.
22- Tashakkori A & Teddlie C (2003) Handbook of mixed methods in social and behavioural research. California: Sage publications.
23- Greene J, Caracelli V & Graham W (1989) Toward a Conceptual Framework for Mixed-Method Evaluation Designs. Educational Evaluation and Policy Analysis 11: 255-274.
24- Johnson R & Onwuegbuzie A (2004) Mixed methods research: A research paradigm whose time has come. Educational Researcher 33: 14-26.
25- Keeney S, Hasson F & McKenna H (2001) A critical review of the Delphi technique as a research methodology for nursing. International Journal of Nursing Studies 38: 195-200.
26- Polit D & Sherman R (1990) Statistical power in nursing research. Nursing Research 39(6): 365-369.
27- Beyea S & Nicoll L (1999) A survey does not equal research. Association of Operating Room Nurses Journal 69(1): 263-264.
28- Richardson A (1994) The health diary: An examination of its use as a data collection method. Journal of advanced nursing 19(4): 782-791.
29- Shadish W, Cook T & Campbell D (2002) Experimental and Quasi-experimental designs for generalized causal inference. Boston: Houghton-Mifflin.
30- Kenny D, Kashy D & Bolger N (1998) Data analysis in social psychology. In The Handbook of Social Psychology (Gilbert D., Fiske S. & Lindzey G. ed. Vol. 1, 4th ed.), Boston, MA: McGraw-Hill.
31- Field A (2009) Discovering statistics using SPSS (3rd ed.). Los Angeles: SAGE.
32- Pallant J (2007) SPSS survival manual (3rd ed.). Crows Nest, NSW: AllenandUnwin.
33- Smith B, Zehle K, Bauman A, Chau J, Hawkshaw B, Frost S & Thomas M (2006) Quantitative methods used in Australian health promotion research: a review of publications from 1992-2002. Health promotion Journal of Australia 17: 32-36
34- Crichton N (2001) Principles of statistical analysis in nursing and healthcare research. Nurse researcher 9: 4-16.
35- Jakobsson U (2004) Statistical presentation and analysis of ordinal data in nursing research. Scandinavian Journal of Caring Sciences 18: 437-440.
36- Hoeing J & Heisey D (2001) The Abuse of Power: The pervasive fallacy of power calculations for data analysis. The American Statistician 55(1): 19-24.
37- Hirt E & Markman K (1995) Multiple explanation: A consider-an-alternative strategy for debiasing judgments. Journal of Personality and Social Psychology 69(6): 1069-1086.
38- Warmbrod J (1970) Interpreting evaluation findings. Rockville: National Institute of Justice, NCJRS paper reproduction. NCJ 015379.
39- Guyatt G, Osoba D, Wu A, Wyrwich K, Norman G & The Clinical Significance Consensus Meeting Group (2002) Methods to Explain the Clinical Significance of Health Status Measures. Mayo Clinic Proceedings 77: 371-383.
40- Bradbury-Jones C (2007) Enhancing rigour in qualitative health research: exploring subjectivity through Peshkin's I's. Journal of Advanced Nursing 59(3): 290-298.
41- Cutcliffe J & McKenna H (1999) Establishing the credibility of qualitative research findings: The plot thickens. Journal of Advanced Nursing 30(2): 374-380.
42- Davies D & Dodd J (2002) Qualitative Research and the Question of Rigor. Qualitative Health Research 12: 279-289.
43- Sandelowski M (1986) The problem of rigor in qualitative research. Advances in Nursing Science 8(3): 27-37.
44- Drew N (2006) Bridging the Distance Between the Objectivism of Research and the Subjectivity of the Researcher. Advances in Nursing Science 29(2): 181-191.
45- Maxwell J (2004) Using Qualitative Methods for Causal Explanation. Field Methods 16: 243-264.
46- Meadows-Oliver M (2009) Does Qualitative Research Have a Place in Evidence based Nursing Practice? Journal of Pediatric Health Care 23(5): 352- -354.
47- Monti E & Tingen M (1999) Multiple Paradigms of Nursing Science. Advances in Nursing Science 21(4): 64-80.
48- Onwuegbuzie A & Leech N (2005) Taking the ''Q'' Out of Research: Teaching Research Methodology Courses Without the Divide Between Quantitative and Qualitative Paradigms. Quality and Quantity 39: 267-296.
49- De Leeuw E (2005) To Mix or Not to Mix Data Collection Modes in Surveys. Journal of Official Statistics 21(2): 233-255.