The first part of this dissertation evaluates the effectiveness of food assistance by gathering evidence from developing countries and the United States. The first essay applies a multi-market model to three developing countries and simulate recent spikes in staple prices and food aid impacts. Results indicate that higher food prices would result in reduction in household real income and deterioration of household welfare. Food aid in the form of cash transfers targeted at low-income groups could improve household real income of the target group after world price shocks and, partially or completely, offset the negative impacts of higher food prices. The impact of cash transfer on untargeted groups is ambiguous. It is likely to be positive for households that are net producers of the commodities that have increased production and prices under cash transfer and the production surplus is sufficiently large. The second essay focuses on the Food Stamp Program (FSP), a cornerstone of food assistance safety net efforts in the U.S. to reduce household food insecurity, particularly among children. The essay examines the dynamic relationship between FSP participation and child food security using monthly measures. Empirical estimates using the Panel Study of Income Dynamics demonstrate that child food security declines in the months immediately prior to FSP entrance, but then partially recovers following program entrance. These dynamic FSP effects are masked when annual measures are employed. The third paper of this dissertation studies the potential impacts of patient's widespread use of online health information. In particular, the essay employs a principal agent model and focuses on the quality of online health information. The model shows that when the quality of health information improves, since medical consultations become more efficient and less costly, a higher effort will be induced or contracted from the physician. Diagnosis becomes more accurate, because physicians will try exert more effort in diagnosing patients and patients will suffer less loss from their illnesses. / Ph. D.

The purpose of the study was to examine the relationships between types of health information tasks and the Internet information search processes of Chinese graduate students at the University of North Texas. the participants' Internet information search processes were examined by looking at the source used to start the search, language selection, use of online translation tools, and time spent. in a computer classroom, 45 Chinese graduate students searched the Internet and completed three health information search tasks: factual task, interpretative task, and exploratory task. Data of the Chinese graduate students’ health information search processes were gathered from Web browser history files, answer sheets, and questionnaires. Parametric and non-parametric statistical analyses were conducted to test the relationships between the types of tasks and variables identified in the search process. Results showed that task types only had a statistically significant impact on the time spent. for the three tasks, the majority of Chinese graduate students used search engines as major sources for the search starting point, utilized English as the primary language, and did not use online translation tools. the participants also reported difficulties in locating relevant answers and recommended ways to be assisted in the future when searching the Internet for health information. the study provided an understanding of Chinese graduate students' health information seeking behavior with an aim to enrich health information user studies. the results of this study contribute to the areas of academic library services, multilingual health information system design, and task-based health information searching.

Yes / Objective To assess the quality, readability and coverage of website information about herbal remedies for menopausal symptoms. Study design A purposive sample of commercial and non-commercial websites was assessed for quality (DISCERN), readability (SMOG) and information coverage. Main outcome measures Non-parametric and parametric tests were used to explain the variability of these factors across types of websites and to assess associations between website quality and information coverage. Results 39 sites were assessed. Median quality and information coverage scores were 44/80 and 11/30 respectively. The median readability score was 18.7, similar to UK broadsheets. Commercial websites scored significantly lower on quality (p = 0.014), but there were no statistical differences for information coverage or readability. There was a significant positive correlation between information quality and coverage scores irrespective of website provider (r = 0.69, p < 0.001, n = 39). Conclusion Overall website quality and information coverage are poor and the required reading level high.