What I am working on - psychometry and Resilience
Hi,
I know I don't see many people make a detailed post on what they are doing. I am not sure if its because of IPR or what. Anyway I have decided to do a detailed post. about my research. Anyone who has any suggestions please email me.
Concept:
I began thinking about Resilience. Everyone meets hard times, how come some people are able to Bounce Back easier than others? Is this a personality trait? Does this have to do with belief systems? Of course this is not the only research in the subject. Salvatore Maddi and Suzanne C. Kobasa pioneered the concept of Hardiness in 1979. Then there is the CD-RISC (Connor Davidson Scale of Resilience). Resilience seemed to be a combination of (1) social support (2) actually, how often does one fall sick.......... but what I was interested in is the EXISTENTIAL dimension. As Viktor Frankl says....."If you have a why to live, you can put up with any HOW" This resonated with a lot of my personal feelings too. Some of the tests had an existential dimension some did not. I decided to focus on the CD RISC as a base. I took a close look at the CD-Risc and I found that most journal publications in the matter is in journals of nursing. This meant that it had not been used outside the context of Health Care. While Salvatore's Hardiness was already into TRAINING people to be Hardy. Living in India, the idea of SPIRITUALITY is strong. Hence I tied up spirituality to the idea of a higher purpose, sense making rather than to any metaphysical idea of GOD, etc. I also felt there are differences in terms of culture. For example in a western culture people attribute success to Internal Locus of Control while in India we would attribute success to external factors.
I looked at existing resilience scales and created a list of items on my own and then asked my associate to do an item analysis to see if i can freeze on the list of items in the test. Can I now say this is an instrument I can use? Of course once I have a valid and reliable i want to use it in additional research.
Here is the response from my associate....
Item analysis - using cronbach alpha as indicator (internal consistency - measuring single concept) at least .7 is needed.
1. With all 26 items, the cronbach alpha (CA) is .431
Since the item analysis also gives the CA if each item is omitted, those which with higher omitted item CA were removed.
2. first level removal - J , Q and H --> CA= .554
3. second level removal - K and O ---> CA= .62
4. third level removal - N, F and C ---> CA=.64
i stopped at that, so now with 18 items it is .6412.
Factor analysis
roughly throws up 3 factors
Now my questions are, as follows.
- Should I go for removal of items as mentioned in step 2-4? If I do will I have an acceptable test?
- Should I abandon my test all together and consider making a fresh one... perhaps I had not FRAMED my items properly?
- Is the analysis appropriate to the situation? Is Cronbach alpha the right analysis for this?
- Can I go for Factor Analysis before my test itself is stable?
If you need my test items to answer my questions, please email me. mathaifenn at gmail dot com