Multiple Imputation and Propensity Scores to Examine the Effects of Improvement Level of Psychological Self-Sufficiency on Economic Self-Sufficiency/Employment Status Among Low-Income Job-Seekers

Background:This study examines the effects of psychological self-sufficiency (PSS) on low-income jobseekers’ economic self-sufficiency (ESS) by using a propensity score matching (PSM) approach. PSS has been studied as one of the crucial factors in terms of improving economic well-being (Hong, Choi, & Key, 2018). PSS is defined as a construct that represents a process of moving from perceived employment barriers to employment hope . Transforming Impossible into Possible (TIP) is an evidence-informed intervention model in workforce development that targets improvement in PSS to help low-income jobseekers achieve ESS (Hong, 2016). While the development of TIP was based on the PSS evidence and theory, no research has yet to test the effectiveness of TIP using a control group design. In order to fill this gap in the literature, this study simulates a randomized controlled trial (RCT) by using a multiple imputation of missing data and propensity score matching (PSM) techniques.

Methods: PSM helps create a model that allows comparison of test effects by using a statistical randomization. Thus, we analyzed the longitudinal data using multiple imputations and PSM analysis to identify the level of PSS over time among low-income jobseekers. Employment Hope scale (EHS) (Hong, Choi, & Polanin, 2014) and Perceived Employment Barrier scale (PEBS) (Hong, Joshua. Key, & Choi, 2014) were used to measure participants’ PSS. and the difference score between EHS and PEBS was used to measure PSS. The WEN Self-Sufficiency Scale (Gowdy & Pearlmutter, 1993) was used to measure ESS. A total of 831 individuals from data collected between 2011 and 2016 were included in the sample. To estimate the propensity score, logistic regression was conducted using PSS as the dependent variable. The PSS values were grouped into one with equal or higher than the mean of PSS difference (Time 2 – Time1) and the other with PSS scores lower than its mean. Age, marital status, level of education, house income, gender, and race were used as control variables. Then we examined the effects of PSSon ESS difference using a multiple regression analysis.

Results: Participants who have equal or higher than the mean PSS difference were matched with participants who have lower than the mean of PSS difference, a total of 144 matched pairs were calculated. Multiple regression results using ESS as the dependent variable showed that of the higher PSS group compared to its reference group had a significantly higher ESS difference score (t = 2.86, p < .01) controlling for employment hope at Time 1, perceived employment barriers at Time 1, and other demographic variables.

Implications: The results indicated thathaving –a higher level of PSS—enhancing employment hope and overcoming perceived barriers—in comparison to having a lower level of PSS is associated with  moving forward toward successful ESS outcomes. Social workers should be aware of the importance of PSS and how it could play a key role in strengthening job readiness training (JRT) programs. Particularly, TIP could emerge as the core integrative intervention model that enhances PSS and thereby ESS and other JRT outcomes.