Dear researchers,
To estimate stability, i calculated ICC. The sample size is 221, the tool included twenty items that consist of two sub scales. The tool scored on a six-point Likert scale.
ICC estimated .02 and .12.
What are ICCS low level?
Thanking you in anticipation.
Hamid
It means the intraclass correlation is low: i.e., it suggests that the measures are not very reliable (i.e., responses to the same item by different participants are not very similar).
ICC suggest the correlation between items (For example: The response in one item is correlated with response in another items). A value closer to 0 would suggest that your items are not correlated with each other. It means the set of items can not be organized in a group.
These values are too low. Are you sure you are doing it correctly? What actually are you trying to do?
Dear Dr Mehmet Sinan Iyisoy
Yes. the sample is real. the sample size is 221 veterans.
What is your aim in calculating ICC? Are you calculating ICC for items? Could you be more explicit?
Aim of our study is determination of validity and reliability of the X scale. The X scale includes 20 items which are scored on a five-point Likert scale from 1 (Completely disagree) to 5 (Completely agree). The reliability of the X scale was first assessed through evaluating its internal consistency and calculating Cronbach’s alpha. Next, we calculated ICC.
Is my explanations enough?
Sincerely Yours
I needed your explanations in order to make sure.
ICC could be used for reliability if you have repeated results of the scale which I suppose is not true for your case. Otherwise you could try inter item reliability statistics like split-half reliability in addition to Cronbach's alpha not ICC.
I was confused as well as my understanding is that ICCs are used when you have two sets of observations to compare, e.g., the agreement in scores assigned by two different raters, the ratings given by twins.
Here, as noted, you could calculate internal consistency and split half reliability. I assume you have examined each subscale separately, as my assumption would be that internal consistency etc would be lower if examining all 20 items.
With 221 participants, you could also try a factor analysis to confirm that the two subscales represent distinct (if related) factors.A larger sample would be better to arrive at a more robust factor solution but you could take a look.
Dear Dr. Hamid Sharif Nia
I think ICC can be used when quantitative measurements. So it may be low.
If your aim is determination of validity and reliability of the X scale, you should be use Cronbach alpha and Split Half method. In my opinion you have the adequate sample size.
The most common internal consistency measure is Cronbach's alpha, which is usually interpreted as the mean of all possible split-half coefficients.
The split-half method assesses the internal consistency of a test. This halves reliability estimate is then stepped up to the full test length using the Spearman–Brown correlation. İf any items on separate halves of a test which have a low correlation (e.g. r = .25) should either be removed or re-written.
The test-retest method assesses the external consistency of a test. It measures the stability of a test over time. It measures the stability of a test over time.
best regards.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis