Return to search

Statistical analysis of pre-employment predictive indexing within the farm credit system

Master of Agribusiness / Department of Agricultural Economics / Allen M. Featherstone / This thesis analyzes the hiring and selection processes of five Farm Credit Services
(FCS) Associations within U.S. AgBank to determine the effectiveness of potential
employee testing and profiling practices as a predictor of success (defined as tenure and
retention) within the organization. The data provided by the five FCS Associations were
used to analyze whether that the results are a successful tool in predicting the success of a
potential employee.
Firm managers are acutely aware of the high cost of onboarding a new employee
regardless of the industry in which the firm operates. Since employee training and
education often takes months, and in some cases, years, it is critical that organizations
select qualified, driven, and success oriented employees so that they can minimize the cost
of hiring of new employees. To select the best candidates, many firms use personality
profiling examinations to determine the candidate’s fit, not only for the job, but also for the
company culture. Analyzing past results can assist managers in evaluating the outcomes of
the time and cost spent seeking the best employee possible.
Analysis was conducted by estimating a binomial logistic regression model using
the test scores for loan officer hires from five Farm Credit Associations for the time period
of 1999-2009. Each of the examined character traits was an independent variable, along
with variables for gender and whether the candidate was a recommended-hire. The
dependent variable is whether the employee is still employed with the Farm Credit
Association. Results show that while some of the independent variables are statistically
significant in predicting the success of an employee, others are not. The implications therein justify the value of the predictive index as an asset to hiring managers, and also
provides direction on which traits are most highly correlated with one another and with the
overall composite score.

Identiferoai:union.ndltd.org:KSU/oai:krex.k-state.edu:2097/14046
Date January 1900
CreatorsUlrich, Timothy Creed
PublisherKansas State University
Source SetsK-State Research Exchange
Languageen_US
Detected LanguageEnglish
TypeThesis

Page generated in 0.0023 seconds