Survey (or correlational) method
The survey method is commonly used to identify the naturally occurring patterning of variables in the ‘real world’ rather than to explain those patterns (though often people want to put an explanatory gloss on them). So to examine whether absence makes the heart grow fonder we could conduct a survey to see if people who are separated from their partners because of travelling away from home (group A) say more positive things about their partners than people who never travel away from home without their partners (group B). This might be an interesting exercise, but the validity of any causal statements made on the basis of such findings would be very limited. For example, if we found from our survey that group A said more positive things about their partners when they were traveling than group B, it would be impossible to demonstrate conclusively that absence was the cause of the difference between groups A and B. In other words, while our survey could show us that absence is associated with a fonder heart, it could not conclusively show that absence actually causes the heart to grow fonder. It is quite possible (odd as it may sound) that the sorts of people who travel away from home without their partners are simply those that like their partners more (so fondness makes the heart go absent). Or perhaps both fondness and absence are caused by something else – for example, social class (i.e. being wealthy makes people both fond and absent). In large part, then, surveys rely on methodologies that identify relationships between variables but do not allow us to make conclusive causal inferences.
The survey method is commonly used to identify the naturally occurring patterning of variables in the ‘real world’ rather than to explain those patterns (though often people want to put an explanatory gloss on them). So to examine whether absence makes the heart grow fonder we could conduct a survey to see if people who are separated from their partners because of travelling away from home (group A) say more positive things about their partners than people who never travel away from home without their partners (group B). This might be an interesting exercise, but the validity of any causal statements made on the basis of such findings would be very limited. For example, if we found from our survey that group A said more positive things about their partners when they were traveling than group B, it would be impossible to demonstrate conclusively that absence was the cause of the difference between groups A and B. In other words, while our survey could show us that absence is associated with a fonder heart, it could not conclusively show that absence actually causes the heart to grow fonder. It is quite possible (odd as it may sound) that the sorts of people who travel away from home without their partners are simply those that like their partners more (so fondness makes the heart go absent). Or perhaps both fondness and absence are caused by something else – for example, social class (i.e. being wealthy makes people both fond and absent). In large part, then, surveys rely on methodologies that identify relationships between variables but do not allow us to make conclusive causal inferences.
No comments:
Post a Comment