Archive | Research Methodology RSS feed for this section

Empirical data is pertinent in validating research – Merits and demerits in relation to the various methods of data collection

1 Feb

By Ng Boon Ka 2008




According to the Oxford English Dictionary (2nd Edition, 1989), empiric is derived from the ancient Greek for experience, έμπειρία, which is ultimately derived from έυ in + περα trial, experiment. Therefore, empirical data is information that is derived from the trials and errors of experience. In this way, the empirical method is similar to the experimental method. However, an essential difference is that in an experiment the different ‘trials’ are strictly manipulated so that an inference can be made as to causation of the observed change that results. This contrasts with the empirical method of aggregating naturally occurring data.[1]

Notably, strict empiricists derive their rules of practice entirely from experience, to the exclusion of philosophical theory. Notwithstanding the disagreement of this author with such approach, the importance of an underlying philosophy will not be discussed in this paper due to the limit of space and time. For the purpose of this paper, this author proposes to discuss the matter at hand in an avant-garde style which seeks to unravel the merits and demerits of empirical data in validating research by reference to various data collection methods. A bold approach shall be taken by going beyond the confinement of traditional data collection methods, namely, questionnaires, interviews and observations, into structured and combined methods such as forecasting tools. This introduction shall be followed by different research areas (physics, metaphysics and divinity) which will determine the use of various kinds of collection tools. The kernel of the discussion shall revolve around core concepts like accuracy, predictability and practicality. Last but not least, the simplicity and clarity of thoughts is captured by a flow chart given in the appendix.


While physics is the science concerned with the study of physical objects and substances, and of natural forces by employing empirical methods, metaphysics, on the other hand, is the part of philosophy that is concerned with trying to understand and describe the nature of truth, life, and reality. A higher level of ‘divine data’, which can be both visible and hidden from mankind, is beyond physics and metaphysics. Hence, it is obvious that only the first category may produce empirical data and not the latter two.

As pointed by Uma Sekaran, though moods, feelings, and attitudes can be guessed by observing facial expressions and other nonverbal behaviors, the cognitive thought processes of individuals cannot be captured.[2] Howbeit, Uma qualified this by introducing projective methods where certain ideas and thoughts that cannot be easily verbalised or that remain at the unconscious levels in the respondents’ minds can usually be brought to the surface though motivational research.[3]

In this regard, while some stretched psychological testing tools, both electronically and traditionally, to test intuition and inspiration, which may be subjective and abstract, it is nigh to impossibility to test divine revelation. In the Islamic context, the Quranic revelations were always accompanied by certain psycho-physiological changes that could easily be perceived by those present with the Prophet S.A.W. As for the revelations themselves, they occurred according to definite measures and in varying time intervals in a way clearly neglectful of the personal state of the person who was receiving them. Put differently, those revelations were taking place irrespective of the Prophet’s grief and sufferings or wishes and aspirations.[4] Hence, Muslims must accept the incident of such Quranic revelation without an iota of doubt and empiricism. Any empirical method which claims to test the revelation would not validate such benign research of the Quranic origin.

This author is of the view that the compilation of the Quranic ayat in the early ages of Islam as a collection of data encompassed both empirical and non-empirical methodologies. It was empirical because of the direct observation by the Prophet’s sahabat and non-empirical as they believed in the words of Allah as long as they were authentically recorded.

The research on the nature of truth, life, and reality can be creatively seen in the context of the booming growth of Islamic finance. The survey on the use of Islamic financial products and services by customers either in the form of interview or questionnaire is a good empirical method to gauge the confidence and satisfaction level of customers. Seen in the light of communicational clarity, personal interviews may enable the identification of nonverbal cues and the building of rapport, thus, affording more insights to the subject matter. The questions may be structured in a systematic manner (e.g. Longitudinal scale or Likert scale in different products) which facilitates the supply of answers by the respondents. These methods had been also improved by the use of computer-assisted interviewing (CAT) which enhances the relevance of empirical data in validating research.

Nonetheless, such empirical data may not be able to identify the actual reasons of the use of Islamic financial products and services in the first place as some may wish to invest in such products regardless of its validity in the Shariah as long as it provides satisfactory returns. In other words, although interviews and questionnaire may churn out empirical data in relation to the confidence and satisfaction level of the customers (which may be expressed); it may not be deep enough to furnish wholesome understanding of its genuineness (which may be hidden or distorted).


Empirical data obtained through observational studies are generally reliable, rich and free from respondent bias. This resonate the ‘first hand theory’ as opposed to hearsay evidence in which the former underlines the importance to appreciate matters first hand rather than relying on transmission of data without credibility and transparency. Historically, in the ninth century, it was amazing that al-Razi hung raw meat at different places to find the locality with the slowest rate of decay which was later recommended for the building of the new Audidi hospital in Baghdad. This experiment by contact with nature has motivated mankind to understand the complexity of God’s universe in an ingenious and diverse manner.[5]

Nevertheless, the aforesaid example on observation may be tainted with extraneous factors which may skew the examination, determination and final result. Likewise, while it is easier to note the effects of environmental influences on specific outcomes through observational studies[6], there are possibilities that empirical data derived thereof may be invalidated by vitiating factors like natural causes which are beyond our control. Biasness that crop into the data collection methods may result in recording errors, interpretational errors, memory lapses, and unreliability of the empirical data.


The use of meteorological tools to examine natural phenomena like the weather not only gives good empirical data of current status quo, but also gives good forecast to a certain extent. Empirical data from forecast is still second to none in most areas of physical sciences. However, a good forecast remains only as a good estimator since certainty of knowledge is still beyond the realm of human’s intellect. In short, the extrapolation of current or past empirical data to predict the outcome in the future may not indicate the whole truth. Assuming a research is done on the date of kiamat (Last Day), is it possible that we use forecasting tools to predict such an incident? Ostensibly, it is an affront to common sense!

Correspondingly, although fortune telling by observation of the palm and face is a type of empirical data upheld by some people, this kind of data derived from such data collection method may not validate a research, for instance, on the success rate of a person in the future. On the same token, biased empirical data can be found in the criminal profiling of certain suspects based on their DNA, cultural background and even religious belief as evidenced by the negative profiling of Muslims as potential terrorist after the Sept. 11 attack. On the flip side, the data collected though clinical (lab) experiment on the food intake of children may be used to gauge their growth rate in the future.


The collection of empirical data largely hinges upon resources such as time, energy and cost of the researcher which might bias the recorded data due to observer fatigue. In the retrieval of preexisting records particularly historical ones, one may doubt the extent of which all records have survived to allow for fair representation of the sample. In spite of that, empirical data is still better than non-empirical data especially in the historically-oriented research. In short, the practicality of the empirical data is matter to be judged according to the area of research that we are interested.


As there is no certainty in knowledge, no single method and data are sufficient to provide a complete comprehension of an area of research. Empirical data from various data collection methods is the most useful source for contemporary research compared to futuristic and historical research. In both the latter cases, there are elements of uncertainty and subjectivity. In essence, it is contended that a research which adopts a multi-method and multi-data approach (the triangulation of methods), may yield more constructive outcomes. Thus, the axiom ‘do not put all your eggs into one basket’ rings a bell!

[2] Uma Sekaran (2003), Research Methods For Business – A Skill-Building Approach, John Wiley & Sons, Inc., 4th Ed, pg. 254.

[3] Techniques such as word associations, sentence completion, thematic apperception tests (TAT), inkblot tests, and the like.

[4] Malik Bennabi (2004), The Qur’anic Phenomenon – An Essay of a Theory on the Qur’an, Islamic Book Trust, pg. xxi.

[5] This point was stated in the earlier assignment on Muslim Contribution to Scientific Research – A Renaissance.

[6] Uma Sekaran (2003), Research Methods For Business – A Skill-Building Approach, John Wiley & Sons, Inc., 4th Ed, pg. 253.