Master of Agribusiness / Department of Agricultural Economics / Vincent Amanor-Boadu / Order picking errors have an adverse effect on performance because they contribute to lost time, resources and customer loyalty. Therefore, it is imperative that organizations reduce errors as much as possible. However, organizations cannot effectively reduce errors until they understand the factors that determine and influence them and can isolate the sources of those errors. Product distribution at Frito Lay is very critical in the supply chain activities of the company and understanding and managing the level of errors that occur at the distribution phase of operations is critical for the firm’s long term sustained competitiveness. This study examines Frito Lay’s order filling processes and how order volumes affect the level of errors. The company uses two types of order picking technologies: prepick and bulk order, conventionally also known as pick-to-light and voice-pick technology respectively. The main objectives of the study are: (a) to examine the impact of size of volume processed at the distribution center on errors recorded for each order pick technology and (b) the impact of regional and seasonal differences across Frito Lay’s distribution network. The data pertaining to pre-pick volume, pre-pick error, bulk volume and bulk error were collected for ten consecutive quarters time period ranging from first quarter of 2009 to the second quarter of 2011 and across 16 divisional distribution centers in four regions of the U.S. The data were organized into a panel for analyses using Stata® 12.1. With no a priori foundation for choosing any particular structural equation form, a number of structural equations were estimated and compared to consistency with economic theory and internal consistency. Two different sets of models were estimated: one for each technology. The regression results from the analysis from the pre-pick order picking technology models showed the quadratic model was the “best” model, whereas the linear model turned out to be the “best” structural form for bulk order picking system. This research provides valuable information to management in attempt to address errors in the order fulfillment system. Because errors may be human, and these human errors may emanate from lack of knowledge or poor skills, they can be addressed with training and education. The human errors may also be a result of processes in the plant. These could be addressed by the reconfiguration of processes and educating people about those processes. Finally, the errors may be motivational, leading to poor focus in executing responsibilities. To address these types of errors, management may choose to implement both positive and negative incentives. Positive incentives will provide rewards to employees who meet error reduction targets that are established at the beginning of certain periods. Negative incentives may include penalties for exceeding pre-specified error thresholds. The Frito Lay system would benefit more from this research if the data had included human resource demographic data as well as economic information. It would have allowed the research to estimate the effect of errors on the economic performance of the different distribution centers and help determine the economically optimal level of errors at the different centers.

This thesis covers two separate studies. The first study, chapter 2, was conducted to evaluate whether there are additional economic returns from producing strawberries in the high tunnel compared to the open-field in Southeast Virginia. We develop and compare budgets for eight strawberry cultivars grown in the two environments and sold under three marketing strategies (pre-pick wholesale, pre-pick retail, and U-pick). Almost all cultivars in the high tunnel generated negative net revenues regardless of the marketing strategy. In contrast, net revenues from open-field cultivars were always positive. In the second study, chapter 3, we used a fractional multinomial logit model to estimate the effect of crop revenues, input costs, and fuel ethanol production on the joint acreages planted to eight U.S major crops (barley, corn, cotton, peanuts, rice, sorghum, soybeans, and wheat). We found a positive and statistically significant marginal effect of the expected peanuts' revenue on its acreage share. The expected corn revenue had a negative average marginal effect on soybean acreage share, and the effect of expected wheat revenue was positive on cotton acreage share and negative on rice acreage share. / Master of Science / This thesis covers two separate research studies. The first study, chapter 2}, was conducted to evaluate whether growing strawberries in a simple, low-cost, and passive heat structure known as a "high tunnel" would yield more profit (the difference between total revenues and total costs) compared to growing them outside in an open-field in Southeast Virginia. We estimate and compare differences between total revenues and costs for eight strawberry cultivars grown in the two environments and sold under three marketing strategies. The first two marketing strategies involve growers harvesting strawberries themselves. They can then either retail them (farm stands, farmers' markets, or in a similar setting: pre-pick retail) or sell them in bulk to be retailed by others (pre-pick wholesale). A third marketing strategy involves consumers visiting a farm and picking their own strawberries (U-pick). Almost all cultivars grown in the high tunnel structure resulted in a loss (negative difference between total revenues and costs) regardless of the marketing strategy. In contrast, cultivars produced in the open-field always generated a profit (positive total revenues and costs difference). In the second study, chapter 3, we estimated the effect of crop revenues, input costs, and fuel ethanol production on acreages planted to eight U.S major crops (barley, corn, cotton, peanuts, rice, sorghum, soybeans, and wheat), and we considered the fact that acreages allocated to one crop affect other crops' acreages. We found that increasing the expected revenue of peanuts leads to an increase in its acreage share. Increasing the expected revenue of corn leads to a decrease in soybeans' acreage share. Finally, increasing the expected revenue of wheat leads to an increase in the cotton acreage share and a decrease in the rice acreage share.