I am confident in the validity of some of these instruments. I address construct

validity by operationalizing the variables under study and then creating instruments based

on the opererationalized definitions. I increase face and content validity by having an

expert panel and colleagues help develop the scales and indices and also by referring to

past research which tried to measure similar constructs within different settings. Two of

the indices became problematic during the data collection process. This reduced the

validity of the instruments.

Table 3-1: Cronbach's alpha and item total correlation values for measured indices
Construct Item total Cronbach's Items removed Items
 correlation Alpha* remaining
 index
Behavior: Use 0.33 0.79 (1) crop insurance 9
of TAP (2) mechanical land
 preparation
 (3) pest scouting
 (4) seeds or cultivators that
 are readily available
 commercially
Self- efficacy .49 .88
Subj ective NONE NONE All subjective norm index 1
Norm items except one-item placed
 with normative beliefs index
Normative .46 .76 (1)Family 4
Beliefs** (2)Other farmers and
 neighbors
 (3)Tribal Elders
Traditional .45 .69 (1) I use a small or family 3
Self- Identity business to run my farm.
 (2) I rely on internal
 resources available to me on
 my farm.
 (3) I am thrifty.
 (4) I take few risks.
Modem Self- .47 .88 (1) I rely on external 9
identity resources available to me
 from off my farm
 (2) I rely on high inputs
 (3) I take risks.
 (4) I want to expand the
 acreage on my farm/garden