| Reliability
and Validity
The latest version of Smart
Hiring uses sample data captured
from various methods. The first
one is wherein company received
data for more then 5000 personnel
tested using Smart Hiring
all over India in various companies.
Smart Hiring was also
administered to 454 prospective
MBA students from 12 Management
Institutions located in
Northern, Eastern, Western and Southern
regions of the country. The
sample includes both male and
female students of 21-26 years age
from the prestigious business
education institutions.
It was a purposive sampling. Efforts
were made to include B-Schools
representative of the HR population,
by keeping proportion of each sex,
geographical distribution, socio-economic
status and age distribution equitable.
Read more by downloading the Test
Manual from the link in the
left column.
RELIABILITY
Reliability refers to
the extent to which a test is internally
consistent i.e how accurately
is the test measuring what it is
designed to measure?
The reliability or internal consistency
of the scale was calculated by Split-half
method. The scale was divided into
two parts using odd-even method.
The correlation co-efficient thus
calculated were corrected by Spearman
Brown formula for each dimension
and for the whole test. For a detailed
correlation co-efficient please
download Smart Hiring Test Manual.
Smart Hiring has high
reliability co-efficient for
each of the personality dimension
and statistically highly significant.
The high reliability coefficient
for the total score shows that
total score is a more dependable
index of personality dimensions
than is any score for each
of personality dimensions.
VALIDITY
Validity refers to the
degree to which a test measures
what it purports to measure.
The Online Version of Smart
Hiring has 150 items that
cover 15 normal personality dimensions.
All the test items have
been pruned and refined
by a team of highly qualified
and experienced psychologists.
Based on consensual approach
regarding manifestation constituting
evidence of a specified trait, the
items were selected and placed under
each of the respective personality
dimensions. For each dimension,
finally, there are 10 questions.
The data collected on 454 candidates
from various B-schools was
also used to validate the items
by evaluating the relevance
of test items individually
and Test as a whole; the purpose
being to learn whether answers to
the individual items are, on the
whole, reasonably consistent with
the behavior or personality trends
suggested by the scores. This is
a form of content validity.
See detailed Validity Co-efficient
by downloading Smart Hiring
Test Manual.
Validity of Smart Hiring
was also checked by administering
a standardized Inventory with
36 items taken from published tests.
Scores on inventory as
a criterion were correlated
with scores on Smart Hiring
for 454 subjects. The correlation
co-efficient between the criterion
and Smart Hiring is
616 which is highly significant
at 1% level. Thus Smart Hiring has
high concurrent validity.
The above finding shows that validity
of the new version of Smart Hiring
is quite high. It may be mentioned
that validity and cross validation
of psychological instrument like
Smart Hiring is a continuous process
at Smart Hiring. With increased
data and feedback coming from valued
users, its predictive value is likely
get enhanced. .
|