Key Features of Program Validity and Program Assessment
DEFM 410
American Military University
, Key Features of Program Validity and Program Assessments 2
Introduction
The goal of public policy is to identify issues within the society and provide social
action to improve conditions. One of the many problems we face in society today is that with
a government so large and active, there is no lacking public policy. Steve M. Teles of Johns
Hopkins University coined the term “kludgeocracy” to describe a complex system whose
costs outweigh its deeds (Teles, 2013). There are so many programs that overlap it is a wonder
that programs have any positive impact on society. Consequently, what makes a policy valid
or feasible, and how do we assess policies for these traits?
Program Validity.
First, validity means "that you can show that you are measuring what you intended to
measure' (Pope, 2017). To determine the validity or feasibility of a program, it is necessary to
conduct a program evaluation. Program evaluation consists of using research methods to
investigate the effectiveness of social programs and determine if they align with the social and
political goals they were designed to support (Rossi et al., 2018). A process of program
evaluation includes research tests that have four types of validity criteria: construct,
convergent, concurrent, and discriminant validity (Jung, 2012
Construct validity combines multiple evidence types or data to build an interpretation of
a program (Bagozzi & Phillips, 1991). It is an evidence type constructed of multiple data points
to prove validity, hence construct validity. Convergent validity compares the new measure to an
already existing measure of validity (Jung, 2012). Concurrent validity takes a program or
measure that is already known to be valid and compares a new measure against it to help
determine the validity of the new measure. Convergent validity looks at the agreement of two
separate program assessments, and discriminant validity tests are programs meant to be
unrelated but actually are (Jung, 2012).
DEFM 410
American Military University
, Key Features of Program Validity and Program Assessments 2
Introduction
The goal of public policy is to identify issues within the society and provide social
action to improve conditions. One of the many problems we face in society today is that with
a government so large and active, there is no lacking public policy. Steve M. Teles of Johns
Hopkins University coined the term “kludgeocracy” to describe a complex system whose
costs outweigh its deeds (Teles, 2013). There are so many programs that overlap it is a wonder
that programs have any positive impact on society. Consequently, what makes a policy valid
or feasible, and how do we assess policies for these traits?
Program Validity.
First, validity means "that you can show that you are measuring what you intended to
measure' (Pope, 2017). To determine the validity or feasibility of a program, it is necessary to
conduct a program evaluation. Program evaluation consists of using research methods to
investigate the effectiveness of social programs and determine if they align with the social and
political goals they were designed to support (Rossi et al., 2018). A process of program
evaluation includes research tests that have four types of validity criteria: construct,
convergent, concurrent, and discriminant validity (Jung, 2012
Construct validity combines multiple evidence types or data to build an interpretation of
a program (Bagozzi & Phillips, 1991). It is an evidence type constructed of multiple data points
to prove validity, hence construct validity. Convergent validity compares the new measure to an
already existing measure of validity (Jung, 2012). Concurrent validity takes a program or
measure that is already known to be valid and compares a new measure against it to help
determine the validity of the new measure. Convergent validity looks at the agreement of two
separate program assessments, and discriminant validity tests are programs meant to be
unrelated but actually are (Jung, 2012).
