Provided that the questionnaire (called also instrument) has been validated according to a standardized and well established methodology. The study design is also important. Longitudinal studies are more reliable than the cross sectional studies. Finally, comparison with a matched sample of general population is also important.
I think QOL can be meausured using both quant and qual and indeed I think it is important to capture both. I do sometimes think that factors that impact upon a person's quality of life are so variable that some measures may miss important aspects. I agree that it is important to use validated measures but it is also important to ensure measures capture what the participant considers to be important for their quality of life, rather than providing criteria for them to fit into.
Depends on your aim and your analysis, and of course the measure(s) you want to use. It can provide both. It also depends how you use the measure.
The quality of life questionnaires are attempts to make objective something that is extremely subjective that it is the quality of life and therefore brings a potential risk of error and should be used very carefully. For this reason there are so many different questionnaires available and validated. Its main advantage is to assess the patient's opinion, which may be different from the health and the research team and are useful in studies of therapy, as an improvement of a biological outcome does not always mean a better sense of well-being and health for the patient.
Maleki, I am using the SF12 with a quantitative analysis (SEM) because we have a big sample of the population in Sao Paulo, but if you are testing a small group or a special case, you probably are doing an accurate qualitative analysis. It depends on your objective and sample of your study. In general QoL scales have not opened questions, so, you can easily quantify them.
The score of the QOL-AD (Logsddon et al, 1999, 2002) is a quantitative score (13-52). No cut-off for good or poor quality of life.
Ref. Conde-Sala, J. L, Garre-Olmo, J., Turró-Garriga, O., López-Pousa, S. & Vilalta-Franch, J. (2009). Factors related to perceived quality of life in patients with Alzheimer’s disease: the patient’s perception compared with that of caregivers. International Journal of Geriatric Psychiatry 24, 585-594.
Qulaity of life is a concept and of course is a qualitative variable. But, for simplicity of measuring quality of life, we change this variable to an quantitative sacle. So, most common questionnaires (such as sf-36) measure this variable as a quantitative, and we can calculate the score of questionnaire to assess quality of life of subjects. Of course, with open questions this concept can be measured as a qualitative variable, but analysis and interpretation of data is not simple!
I would like to thank all the responders to my question. As I can see there are some conflict in this concept. All we know while measuring QOL or health related QOL, we are measuring non-numerical data (like having energy, social or sexual life, feeling good etc). But as a need to compare between various people either patients or general population we do convert these data into numerical measures/ scales with the use of standardized questionnaires. So what do we get in the end? Is the score really a numerical data? Let me ask in this way, when somebody has a score of 50 in a study using SF36 questionnaire, does that mean that he/she has a QOL twice as much as a person having a score of 25; as it is in comparing height or weight of people( A man with a weight of 120 kg has a weight doubled as another man with a 60 kg weight)?!
So are we allowed to calculate means for these scores? Can we use other parametric statistical tests for these scores? If yes why?
Dear Iradj,
It is for this reason that I underlined the important role of the longitudinal studies that measures the dynamic character of the evolution of the QOL. In this type of study, the score doesn't have any importance because one measures the improvement or the deterioration of the QOL compared to the baseline. You can have an example while reading my article on the assessment of the QOL before and after liver transplantation [Karam et al. Liver Transplantation 2003]. I also underlined the important role of the assessment of a sample of the general population matched to the studied cohort on some criterias as age, the sex and the social statute. It permits to see how is located the scores of the cohort in relation to "norms" of the general population. Thus, the score itself is not important because it is the "delta" in relation to the reference that counts [Karam and al. Transplantation 2003].
However, the score can be standard in the case of questionnaires validated in several languages as the SF36. The score's calculation of domains of the SF36 is standard and can be compared with those published by other teams. However, the SF36 is only a generic questionnaire that only evaluates general aspects of the QOL. The ideal is a joined administration of the SF36 and the specific standard questionnaire to the pathology (cohorte) that you project to study.
We change the qulaity of life to a numeric variable with interval scale, not ratio scale, So, of course when somebody has a score of 50 in a study using SF36 questionnaire, it does not mean that he/she has a QOL twice as much as a person having a score of 25, because our scale does not have a real zero. We have to consider, comparison of QOL score between groups has an important role.
It's been a long time since grad school, but it would seem to me that any scale that attempts to measure a subjective concept necessarily produces nominal or ordinal data depending on the items in the scale. Endorsing more than one item on the scale may produce categorical data. It might be perceived as ordinal if it's reasonable to add up the number of items endorsed and assume endorsing more than one item implies as an increase in the measured attribute. If a scale could reasonably called ordinal, some might considered it marginally quantitative. But I would argue that reporting the score, the median or percentage is potentially misleading at best since the distance from one score to another is unknown and unmeasurable. I think it is also possible to start with a subjective concept, define it operationally as quantitative and measure it as if it's interval data. However, one might argue that you are no longer talking about the original concept, rather the concept produced by the operational definition, which is likely at best only a factor of the full concept.
QOL is qualitative measurement tool and the Scores will help to know how much the progression of the condition/aspect of being measured.
Quality, itself subjective as quality cannot be measured in an objective manner. We cannot assign a specific number either in terms of percentage as it is 78 or 345 per cent good or any value on interval or ratio scale. It can be given in terms of scores (ranks limited by the researcher to a manageable number for analysis) or in terms of nominal terms.
However, QOL approached by qualitative analysis which may come up a number of themes of satisfaction or dissatisfaction or approached using scores in both cases users of amenities, facilities, infrastructure, security and other services, employment condition and income etc. peoples opinions will manifest in the two approaches. However, scores can be statistically analysed using non-parametric methods
Historically, quality of life is a qualitative variable which normally measured in psycho-social sciences but since psychometric found quality of life measure was quantified into a statistical scale. Since then WHO included world wide researchers were produce instruments to measure Qol quantitatively.