Date of Publication


Document Type


Degree Name

Master of Science (M.S.) in Psychology






Pregnant women -- Health and hygiene, Health behavior -- Forecasting, Pregnant women -- Social conditions, Pregnant women -- Economic conditions, Newborn infants -- Health and hygiene



Physical Description

1 online resource, (147p.)


There is a persistent relationship between socioeconomic status and physical health outcomes found in the literature; however the variables mediating this relationship are many, and ways that they interact with each other are complex. The goal of understanding this relationship is to decrease the disparity in health by socioeconomic status.

This study tested a biopsychosocial model proposed by David Williams (1990) to explain the relationship between socioeconomic status and physical health outcomes. The model included the following latent factors: demographics, socioeconomic status, biomedical risk, medical care, psychosocial variables, and health outcomes. The model was tested through a secondary data analysis.

The 1937 women who participated entered one of six Portland area clinics over a three year period for prenatal care. Data was collected over three time points; two interviewsdone during the pregnancy and birth outcome data, taken from medical records. There were 1134 women with complete data for the purpose of this analysis.

Confirmatory factor analysis was used to verify that the measurement tools fit the measurement model for the latent factors. After dropping four of the 28 measures, the fit was adequate and covariance structure modelling was used to test the structural model proposed by Williams. The fit of the model was adequate, however, only 3.8% of the variance in the outcomes measures was explained by the model, and three of the five paths leading to outcomes were insignificant.

An alternative model with psychosocial variables broken into the two factors of psychosocial resources and behaviors was also tested with similar results. The variance explained in outcomes was 4.3% and the only factors with paths significantly related to outcomes were demographics and biomedical risk.

It is of note in both models that SES was a very powerful predictor of the medical care variable, predicting over 50% of its variance. The psychosocial variable also had 20.7% and 18.8% of its variance explained by the preceding factors in Williams' model and the alternative model respectively.

Since the model did fit the data, it is believed improvements in utility of the model could be seen if a study was designed specifically for testing this model.


In Copyright. URI: This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).


If you are the rightful copyright holder of this dissertation or thesis and wish to have it removed from the Open Access Collection, please submit a request to and include clear identification of the work, preferably with URL.

Persistent Identifier