Dear all,
I want to analayze the inattentiveness level in ADHD patients and compare the current score with the score of 6 months before now. I do not have any score of inattentiveness level of 6 months before now. If I ask the patient or the patients close relative to state or fill out a questionnare about inattentiveness level of 6 months before now( stating their behaviour of 6 months earlier), will the results be reliable? Will the comparison stay out of bias?
Patients' reports on Past Week is already questionable, let alone 6 months...especially when they have ADHD, a condition associated with commonly poor memory. I think you can safely assume that subjective reliability will be very low. If you don't have data from 6 months ago, ask patients the 6-month question anyway and in your report clearly state that the recall is only memory-based and provide reasons for potentially low reliability. If they report overall improvement, esp if they have improved grades to show it (assuming we're talking about students), use that as a part of your analysis.
You measure what you measure. So what you are measuring is how the person sees the past in the light of what has happened since. And, of course, both their view of the past and their view of the present are taking place on the same occasion, so they will have a significant common bias. See further under "Method Bias"
Podsakoff, P., MacKenzie, S., Podsakoff, N. (2012). Sources of Method Bias in Social Science Research and Recommendations on How to Control It Annual Review of Psychology 63(1), 539-569. https://dx.doi.org/10.1146/annurev-psych-120710-100452
There is little point in asking them to fill out a scale twice, once for today and once for six months ago. You could ask them to use a scale for each item that runs from much worse through no change to much better. However, you run the risk that people are likely to tell you what they would like you to hear. If you have spent time and trouble helping them, they don't want to tell you that you wasted your time. This is called courtesy bias.
And – as if this were not enough – there is the question of a null hypothesis. If you did a programme to improve their ADHD, then there is no believable hypothesis that it made no difference. The question is how big the difference should be in order to declare the programme a success. With a cheap, easily-scalable programme, this could be a small difference. But if the programme is expensive in time and/or resources, then the effect has to be much larger.
Ah – education research. Don't you love it!
If you could that by observation of a control sample, that will be staying out of the bias and I think is more reliable .
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Probability
- Probability Theory
- Reliability