Glei and Goldman

Limitations in Glei and Goldman Research
Name
University
How to View Pregnancy-Related Care in Guatemala
The article titled Understanding Ethnic Variation in Pregnancy-related Care in Rural Guatemala was published in the year 2000 by Noreen Goldman and Dana A Glei. The article is facilities, socioeconomic factors and other control variables on the ethnic variations in prenatal care in rural Guatemala (Glei & Goldman, 2000). The researchers wanted to identify the factors that influenced care-related decision making among pregnant women in four rural communities, and how choices varied among the three groups of indigenous women compared to their ladino counterparts. They also sought to confirm and account for previous study findings which showed that indigenous mothers were most likely to depend on midwives while ladino women preferred biomedical care. Using quantitative data obtained by the Guatemalan Survey of Family Health comprising of 2,872 participants aged 18 to 35, the researchers studied trends in sourcing for pregnancy, birth, and postnatal care subject to determined independent and control variables. Through quantitative analysis, the researchers evaluated data depending on two factors (a) the nature of care received; and (b) the location of the birth. The findings demonstrate that there is a significant ethnic difference in how women source pregnancy-related care, with sociocultural factors having the most significant impact in decision-making, and midwives playing a central role in the provision of services.

Though data sourced from the Survey provides and detailed statistics from the target population, the sample is not representative. It only investigates the situation in which indigenous women do not use modern medical care. It does not describe in detail why they do not use it. Also, the use of quantitative methods limit the scientific findings achieved by the study because it prevented researchers from getting more information from the available data. A critique of this article considers the research design conducted by the Survey, and how it limits or enhances the validity and reliability of this research. It will also look at the analysis procedure used in this study and ways the researchers could improve it using qualitative data to address their hypothesis.
The first point is that use of existing data to conduct a study provides researchers with several benefits, but it also limits them to the quality associated with such data, and limits how researchers can use the information and the variables available to them for testing. Because the existing data has relative limitations, it hinders the researchers from studying of the conditions under which women are using modern medicine. For the article of Glei and Goldman, the Guatemalan Survey of Family Health gave them statistics on past data between 1990 and 1995, collected from participant recollections about their choice of pregnancy care. It enabled researchers to overcome some of the sampling and data collection problems faced by previous studies in reaching a representative sample. The model has the potential for bias since in the five years, since the data was collected. The social and cultural statuses of the respondents could have changed, affecting how they make decisions about the care they receive during pregnancy. However, the data gathered by the Survey does not cover these variables and their impact on sourcing care. Glei and Goldman identify some standard errors that could arise from the sampling process used in the Survey, so they estimate variances in the results using STATA and include dummy variables in their multivariate models to make up sample stratification at department levels. Since the researchers were not involved in the initial data collection process, however, they probably overlooked other glitches that could have been important in interpreting certain variables (Cheng & Phillips, 2014). Samples also came from the departments that contained two of the largest communities of indigenous women, excluding the smaller societies living in the most remote areas in the country, of which there are eighteen. The findings obtained from this study will not be representative of the national situation since the variables associated with various people were not explored. Although depending on existing data might have helped to overcome specific challenges posed by the scope of the study, but it introduced issues that put the validity and reliability of the research to question.
The second point is that although the data that was available in the Guatemalan Survey of Family Health (EGSF) permitted the researchers comprehension of the relationship between the pregnancy-related care in Guatemala and ethnicity, it limits the scope and extensibility of the study to the Guatemalan population. This is because the EGSF was primarily based on only four of the 22 departments in Guatemala and it disregards the areas that are most remote and therefore, excludes many of the groups that are indigenous to Guatemala (Glei and Goldman, 2000). Moreover, even if the EGSF had detailed data on the time and distance to facilities and providers, the quality of the care was excluded and this does not account for the choices that are made by the Guatemalan population in the selection of either midwifery services or the use of modernized hospital facilities in the delivery of birth. Because it was not representative of the rural Guatemalan population. Within the context of this study that entailed the study of why ethnic variation exists in pregnancy-related care in Rural Guatemala, there are many factors that contribute towards the characteristics or actions of the population, for example, apart from socioeconomic status, one’s proximity or access to healthcare can be a major factor in pregnancy. Lastly, because the research uses a dummy variable to extrapolate their findings and define a definite outcome, some errors might have been encountered in the process, and because they did not employ any specific measures that reduce bias due to errors, the results of the study were bound to be inaccurate or lean towards inaccuracy. This is because the measurement errors of a dummy variable are usually negatively correlated with the true variable (Pischke, 2007) meaning that as the true value decrease the error increases and vice versa. In this manner, all these limitations that are involved in the study largely affect the generalizability of the results or findings of the research to the whole Guatemalan population.
It was advantageous to research to use archival information because they saved on costs and because the research is inclusive of long periods, it allows a broader observation of outcomes and trends. However, because the research did not use their primary data, there was no way they could tell if it had any errors or not. For instance, if the EGSF data had errors, the same errors were transferred to this research thus making it at risk of being error prone. Moreover, the EGSF archival data is old (Jones, 2010) as it was collected between 1990 and 1995. This means that the study by Glei and Goldman did not account for 5 years’ worth of information that is pregnancy related in the Guatemalan region. This is because there might have been temporary or permanent changes during the five-year period to make significant impacts to the results of the research. This can be partly because there existed a civil war in Guatemala from 1960 to 1996. Since the EGSF data was based within these years of civil war, its end might have had a significant effect on the normal social classes and the modes of societal characteristics. For instance, the civil war might have been a cause for the women in Guatemalan region to use midwifery services because of fear of meeting militias. Moreover, due to the civil war in the Guatemalan society, biomedical equipment in the hospital, medicines, and even personnel were limited. This might have a bias towards the Guatemalan women to use midwifery services as their modes of delivery. In this manner, the data that was used by Glei and Goldman was biased due to changes that took place in the societal characteristics before, during and after the civil war. To offset this bias, the researchers should use their primary data that is current and void of the effects of the civil war on the Guatemalan society.
In conclusion, the data collection methods employed by the Survey and the analysis employed by the researchers are two key aspects that limited the scope of the research and made it so that the study fell short of what it could have accomplished with the data and methods available. Also, the Survey also provided an unrepresentative sample and did not cover all the issues that this research could have analyzed to address the set hypothesis. It is critical for researchers to assess the opportunities provided by existing data and ensure that they outweigh the limitations they offer before using such sources because, otherwise, they will not achieve their objectives.

References
Cheng, H. G., & Phillips, M. R. (2014). Secondary analysis of existing data: Opportunities and implementation. Shanghai Archives of Psychiatry, 26(6), 371.
Glei, D. A., & Goldman, N. (2000). Understanding ethnic variation in pregnancy-related care in rural Guatemala. Ethnicity & Health, 5(1), 5-22.
Jones, C. (2010). Archival data: Advantages and disadvantages for research in psychology. Social and Personality Psychology Compass, 4(11), 1008-1017.
Pischke, Steve. (2007). Lecture Notes on Measurement Error. http://econ.lse.ac.uk/staff/spischke/ec524/Merr_new.pdf