Microsoft Word research method fm doc

 
Background variables
- In almost every study involving human subjects, background 
variables, such as, age, sex, educational status, monthly family income, marital status and 

---

Research methodology 
36
religion will be included. These background variables are often related to a number of 
independent variables, so that they influence the problem indirectly. Hence they are called 
background variables or background characteristics. 
Confounding variable
- A variable that is associated with the problem and with a possible 
cause of the problem is a potential confounding variable. This type of variable may either 
strengthen or weaken the apparent relationship between the problem and a possible cause. 
Composite variable
- A variable based on two or more other variables may be termed a 
composite variable. Incidence and prevalence rates, sex ratios, and other rates and ratios are 
composite variables, since they are based on separate numerator and denominator 
information. 
I. Operationalising variables by choosing appropriate indicators 
Note that the different values of many of the variables presented above can easily be 
determined. However, for some variables it is sometimes not possible to find meaningful 
categories unless the variables are made operational with one or more precise 
INDICATORS
. Operationalising variables means that you make them ‘measurable'. 
For example:
1. In a study on VCT acceptance, you want to determine the level of knowledge 
concerning HIV in order to find out to what extent the factor ‘poor knowledge’ influences 
willingness to be tested for HIV. The variable ‘level of knowledge’ cannot be measured as 
such. You would need to develop a series of questions to assess a person’s knowledge, 
for example on modes of transmission of HIV and its prevention methods. The answers to 
these questions form an indicator of someone’s knowledge on this issue, which can then 
be categorised. If 10 questions were asked, you might decide that the knowledge of those 
with: 
— 0 to 3 correct answers is poor, 
— 4 to 6 correct answers is reasonable, and 
— 7 to 10 correct answers is good. 

---

Research methodology 
37
When defining variables on the basis of the problem analysis diagram, it is important to 
realise which variables are measurable as such and which ones need indicators. Once 
appropriate indicators have been identified we know exactly what information we are looking 
for. This makes the collection of data as well as the analysis more focused and efficient. 
2. Nutritional status of under-5 year olds is another example of a variable that cannot be 
measured directly and for which you would need to choose appropriate indicators. Widely 
used indicators for nutritional status include weight for age, weight for height, height for age, 
and upper-arm circumference. For the classification of nutritional status, internationally 
accepted categories already exist, which are based on standard growth curves. For the 
indicator weight/age, for example, children are: 
• Well nourished if they are above 80% of the standard 
• Moderately malnourished if they are between 60% and 80% 
• Severely malnourished if they are below 60%