Colleagues:
Several doctoral students have asked for the psychometric results for the new LMX-SLX. Below are the results. If your doctoral student wishes to use this scale without charge, please ask them to email a short proposal to me.
6 Item (Short Form)
CITC
M
SD
Max IIF
How satisfied is your (colleague, supervisor, or subordinate) with your work?
.84
3.52
1.12
2.92
My (colleague, supervisor, subordinate) would help me with my job problem?
.90
3.61
1.16
8.06
My (colleague, supervisor, subordinate) has confidence in my ideas?
.89
3.59
1.09
8.44
My (colleague, supervisor, subordinate) has trust that I would carry my workload?
.88
3.64
5.38
My (colleague, supervisor, subordinate) has respect for my capabilities?
3.65
5.44
I have an excellent working relationship with my (colleague, supervisor, subordinate)?
.85
1.08
3.55
Note. --CITC = corrected item total correlation. Max IIF = maximum value of information. N=486. Results of modified parallel analysis = unidimensional scale. Internal consistency of responses = alpha = 0.95. Slight negative skewness = do not interpret both mean and variance of LMX-SLX (results will be uninterruptable for variance). This skewness produces spuriously high correlations between LMX mean and variance and variance within groups, and correlations of LMX variance within groups as a predictor are spurious. Linear agreement correlation between supervisor and subordinate reports on same scales was .68 (p<.001) indicating good agreement across the range. Agreement between peers also correlated .68 (p<.001). Moreover, agreement for high LMX dyad is significantly better. LMX differentiation as measured by LMX variance with groups is an artifact of this skewness.
Journal of Business Psychology (2011), 26, 347-357.
New Multinational Network Sharing, 2007, 129-156.
George Graen
/jag