In order to examine the problems affecting Saudi-Arabian students in the United States of America, it will first be necessary to narrow and reframe the broad, general research questions provided elsewhere in this work so that an appropriate strategy with which to approach methodology can be devised. It becomes apparent, upon reviewing the literature, that much of the past thought on the topic of what troubles Saudi Arabian students traveling abroad to attend United States universities, whether through the King Abdullah Foreign Scholarship Program or not, has fallen into two camps. One school of thought seems to be that, in effect, Saudi Arabian students bring their problems with them to the United States, that they struggle primarily with their cultural beliefs and with the wide variety of religious teachings that fall broadly under the umbrella of Muslim thought (Abdullah, 2008). To this way of thinking, the problem is not that they have traveled thousands of miles into a foreign culture; rather, it is that to be Saudi Arabian is somehow inherently problematic. At times, the Saudi Higher Education Ministry is mentioned as well. Needless to say, this is not a direction that seems fruitful to take in determining just what variables to investigate here.
The other school of thought, though, seems equally troublesome; rather than lay blame at the feet of Saudi Arabian culture, some would have it that students and teachers in the United States make education difficult for Saudi Arabian students in universities in the United States by rejecting those students or by projecting some sort of fear of terrorism onto anyone who appears to have a connection to the Middle East. Mention of the notable events of September 11, 2001, is not entirely uncommon in those writing from such a viewpoint (Denman & Hilal, 2011), and yet, considering that over twelve years have now elapsed since that date as of the time of this writing, it is safe to say that the specter of anti-Muslim sentiment, if such sentiment exists, can no longer be blithely attributed to one day in history during which an eighteen-year-old college student today would have been only six years old at any rate and thus incapable of forming terribly strong opinions on political, global issues. Also part of this perspective is the idea that institutions in the United States have not done enough to set programs for Saudi Arabian students. However, going forward, as all things must, this paper leaves behind such simplistic approaches to adopt a methodology more suited to a mature, adult discussion of the experience of Saudi Arabian students in universities in the United States.
Going into this study, some notable assumptions have already been made but not yet expressly articulated. Some are valid, but others will warrant experimental exploration that will lead directly to the creation of a methodology for this study. To begin with, an underlying assumption is that Saudi Arabian students are an identifiable group with distinct characteristics that makes them a sensible choice of the population to sample. This will be considered valid going forward, mostly out of necessity as the generation of methodology on this topic will be nearly impossible otherwise. Another assumption is that Saudi Arabian students have “problems” when attending universities in the United States. This is less obvious; the results of multiple studies, taken as a whole, seem rather mixed on the topic as seen in the literature review. It may be the case that the motivation to look for problems exists in part due to the perception that vast sums of money have been thrown at Saudi Arabian students via the King Abdullah Foreign Scholarship Program in a rather sudden fashion that implies a lack of forethought, though of course, the realities were more complex than such a simple summation implies (Clary & Karlin, 2011; Krieger, 2007). Still, it seems that what would actually serve as a novel approach in this field is to ask the Saudi Arabian students themselves whether they have greater troubles than the average college student in the United States, or even whether they have greater troubles than the average foreign student in the United States, in their own perceptions.
Naturally, there must be a control group to balance out the Saudi Arabian students’ perspective and input into the study. Rosenbaum and Rubin (1985) describe a very similar setup to the one in the process of being proposed in this study, and though their use of propensity score matching is less applicable, the basic notion of control groups as being relevant to this type of non-experimental study can be noted. This control group will consist of United States-born students, as identified simply by the qualifying question, “Do you consider your origin to be primarily from the United States of America?” which thus allows the various nuances of identity that might arise from the situation of a student having ties to multiple countries to be resolved via simple self-identification. Note that here and elsewhere, the term “United States-born” will be used as a phrase that describes such students, as the term “American” could mean anything from Canadian to Peruvian, and thus is too broad to have much scientific value. As with the identity question for United States-born students, it then makes the most sense to use the corresponding question to identify Saudi Arabian students, simply replacing the words “United States of America” with the words “Saudi Arabia” in the question. That such categories are what the study seeks in its participants will be made obvious on the recruitment posters, as discussed below. Having determined in what ways the two groups will be identified, it is now necessary to delve deeper into the type of data gathered.
In order to examine the consonance or dissonance between Saudi Arabian students’ perceptions of their position in universities in the United States and the perceptions of United States-born students, the most appropriate approach is one that is mixed with regard to qualitative and quantitative aspects.
Though the temptation may be to follow the trend of including as many variables as possible, now that the advent of the modern technological age has made it possible for computers to process such data on reasonable time scales, in the end, the troubles of a multivariate approach, from a statistical point of view, are not worth the possibility of gaining additional data in one fell swoop, as it were. Problems already inherently arise from the question of whether survey information collected with tools such as the Likert scale can be truly considered ratio data as opposed to ordinal data, and Fortheringham and Wong (1991) examine an area with similarly messy boundaries to yield depressing results: “ . . . multivariate statistical analysis . . . is, therefore, a much greater problem than . . . univariate or bivariate analysis. The results of this analysis are rather depressing in that they provide strong evidence of the unreliability of any multivariate analysis . . .” (p. 1025). Thus, as the survey nature of the study approach is unavoidable for the research questions being asked, the least that can be done is to mitigate statistical skepticism by keeping the number of variables to one independent variable and one dependent variable, the cleanest scientific approach possible.
The independent variable is national origin, as divided into two nominal categories: those from the United States and those from Saudi Arabia. Though it may be tempting to generate a second independent variable, such as age, income, or the student’s level of college preparedness going in, in fact, this only adds a layer of needless complication and veritably even obfuscation. One independent variable is enough to generate sufficient robustness for this study by means of investigating the topic with greater depth, limiting breadth in order to delve more deeply into one area. The dependent variable is the survey results, and here, there is room for a little more diversity in the topic explored. Student perceptions of their own academic progress and the progress of the other group (i.e., Saudi Arabian students’ opinions of United States-born students will be examined and vice versa) will be examined, as will the idea of there being “problems” arising from any one of a number of areas commonly proposed as being relevant to Saudi Arabian students studying in the United States, as seen in the literature review. Such areas include the level of formality, gender mixing, language barriers, other students’ attitudes, and allowance for religious practices. If being a Saudi Arabian student rather than a United States-born student truly affects the degree of “problems” experienced, then to input that as an independent variable should yield some sort of difference in the output of the survey results as the dependent variable.
The simplest tactic is the best one when it comes to the recruitment of participants. Posters will be placed around the university asking for volunteers and providing a small stipend for their time, depending upon the final budget of the study. The posters will ask for eighteen-to-twenty-two-year-old college students of either United States or Saudi Arabian origin. The age range is specified because Saudi Arabian students studying in the United States are unlikely to be returning students, and thus ensuring that the ages of the two groups are more or less the same—i.e., “college-age”—reduces the effects of a possible confound to the study. As well, the students must be undergraduates, again to keep the study participant recruitment process simple. Ideally, if a translator can be found, the poster will be duplicated in a variety of Arabic that will be understood by Saudi Arabian students. Both posters, in this instance, would as for both United States-born and Saudi Arabian students, for the sake of symmetry and due to the fact that it is not impossible there may be friendship ties between the two groups of students that would allow, for example, a Saudi Arabian student noticing the Arabic poster to bring the study to the attention of a United States-born friend. Indeed, it is hoped that there are friendship ties between the two groups, for the lack of existence of such relationships would be disheartening to the view of the university as a truly multicultural campus indeed, and such friendships would not constitute a significant confound to the study.
Ideally, the sample size would be at least thirty for each group, following the old statistical rule of thumb that n must be greater than or equal to thirty for a trustworthy analysis. If more participants are garnered, the only limits on the potential size of the study are time and space. The survey can be given in any reasonably quiet and private area, such as a library study room, though it would be more ideal to be able to use a departmental office to convey the legitimacy of the study. Thus expanding the study is quite an easy task that consists primarily of making more copies of the survey and inputting their data into the computer to be analyzed.
Next comes a rather sticky issue that perhaps poses ethical challenges, for, in order for the study to have maximum validity, the two samples really ought to be matched with regard to gender. Saudi Arabian students are more commonly men than women due to cultural reasons, but if the other sample is recruited from the more mixed United States-born population of university students, it will likely not match the Saudi Arabian sample very well in terms of gender. It is proposed here that the best solution is indeed to match the participants. “Over the past 25 years, evaluators of social programs have searched for nonexperimental methods that can substitute effectively for experimental ones. Recently, the spotlight has focused on one method, propensity score matching . . . as the suggested approach for . . . education programs,” as Peikes, Moreno, and Orzol (2008) put it. Whereas this study need not go as deep as full-on propensity score matching, the simpler form of ensuring a more or less even ratio—say, within five percent—of males to females in the two groups seems expedient and necessary, yet this will likely mean excluding some female United States-born students from the study, inviting the accusation of sexism. The best thing to do at this point in time is to submit this study as-is to the Institutional Review Board and see what type of response is thus garnered, then to revise as necessary.
The study will use a seven-point Likert-type scale to administer a survey with subjective statements that look to examine students’ beliefs about their own performance in school and that of the other group. Ideally, again with a translator, the study can be administered in both English and Arabic, but otherwise, the participants should all briefly rate their own ability to read and understand English. Responses on the scale below a six agreeing to good English comprehension shall lead to the participant being removed from the study. There as in elsewhere, the choice of a seven-point scale is in truth somewhat arbitrary, but as the literature on the topic seems to find no particular reason for using any certain number, in the end, the path that makes the most sense is the one that is the most intuitive to the researcher in question. Though it may be a rather surprising result, Matell and Jacoby (1971) find that “Results show that (a) there were no specific patterns for reliability or validity coefficients using different number of alternatives; and (b) when the original scales were dichotomized or trichotomized, no significant differences were found in reliability or validity” (p. 657). Albaum (1997), too, touches on the issue, with similar conclusions that no particular number need be favored over another. Thus, as this author has most frequently seen seven-point scales used, that basis alone shall serve as adequate means to point the study in the direction of using a seven-point scale. When the research fails to illuminate any one direction, the person performing the study has no reason not to choose that which will make the interpretation as fluid as possible, even though such matters may ultimately be no more than personal preference.
The questions early in the survey will be simply about academic performance, asking the same question from several angles for a total of around five questions in this category. An example might be: (1) I feel I perform well academically at this school. From there, the survey will move into more general questions about students’ perceptions of their performance as it relates to others, e.g., (2) United States-born students generally perform well academically at this school, or (3) United States-born students generally perform better academically at this school than do foreign-born students. Gradually narrowing the questions, still asking them multiple ways in batteries of about five apiece, the question then becomes: (4) United States-born students generally perform better academically at this school than do Saudi Arabia-born students. Here, note that the language “Saudi Arabian” might be confusingly conflated with Saudi-American, and thus is avoided.
In addition, though it may seem odd at first, the opposite statements must be posed for evaluation on the Likert scale: (5) Students from the United States have greater trouble in universities in the United States than do students from Saudi Arabia, (6) Students from the United States have greater trouble in universities in the United States than do students from foreign countries. Though the answers may be facile “strongly disagree” responses, in fact, asking these converse variants of the previous questions provides this study with greater objective validity and helps erase some of the sense presented earlier that invalid assumptions may have been made. This pair of questions will also serve as a form of validation, for it should not be expected that students studying in their own country would have greater difficulties than foreign students of any kind. If these questions do not average below a two in response, it would be cause for investigation in the discussion and a notable area for future study. It should not be expected that everyone would respond with a one, as some respondents may not be comfortable with putting one or seven answers for any item on the questionnaire, as this implies a degree of certainty not all people experience in their lives. In addition, some may see the possibility that there are certain slight advantages conferred by being an outsider to the culture, such as having a more objective viewpoint. Thus, by asking the opposite questions, this research will confirm that the original questions lie along a valid axis to investigate in the course of this study.
Moving even deeper into the survey, the identified areas of interest must be addressed, namely, the effects of level of formality, gender mixing, language barriers, other students’ attitudes, and allowance for religious practices on students’ academic performance, as perceived by the participants. An example might be: (7) I am not at all distracted by the even mix of men and women on this campus. The fact that the United States-born students would be asked the same questions as the Saudi Arabian students may yield interesting results in this regard; there may be some who are not insignificantly distracted by the same factors that are so commonly cited as troublesome to Saudi Arabian students in United States universities. Finally, if the study wishes to go deeper, questions about the language barrier might be raised, but at this time, it seems the survey will be quite long as stands and needs no additional items.
The results will be analyzed using a chi-square analysis. Lancaster and Seneta (1969) describe the use of such a strategy and imply its appropriateness for the current situation, particularly in the case with one degree of freedom. Thus, this is a valid approach. As this study is in the “soft” sciences, touching on areas of sociology, education, psychology, and anthropology, the appropriate p-value for significance is p < 0.05. The items on the survey will be taken as groups based on the banks of questions in which they were arranged to begin with. Thus, one analysis might be of United States-born versus Saudi Arabian students’ perceptions of Saudi Arabian students’ academic success. This process will be continued throughout the study.
Some extraneous variables not here controlled for, such as the differences in age between eighteen-year-olds and twenty-two-year-olds, or potential income gaps. Neither is the King Abdullah Foreign Scholarship Program status of the Saudi Arabian students taken into account. The researcher leaves it for future studies to tease out any potential differences between groups there.
The next step is for the researcher to use the internet resources available to find, download, print, fill out, and turn in any appropriate forms such that the study can be submitted for the approval of the Institutional Review Board. Having attempted to take into account all potential ethical issues already, it is anticipated that the study will be approved on the first pass.
Abdullah, F. (2008). Combating extremism: A brief overview of Saudi Arabia's approach. Middle East Policy, 15(2), 111.
Albaum, G. (1997). The Likert scale revisited. Journal of Market Research, 39, 331-348.
Clary, C., & Karlin, M. E. (2011). Saudi Arabia's reform gamble. Survival, 53(5), 15-20.
Denman, B. D., & Hilal, K. T. (2011). From barriers to bridges: An investigation on Saudi student mobility (2006–2009). International Review of Education, 57(3-4), 299-318.
Fotheringham, A. S., & Wong, D. W. (1991). The modifiable areal unit problem in multivariate statistical analysis. Environment and Planning A, 23(7), 1025-1044.
Krieger, Z. (2007). Saudi Arabia puts its billions behind Western-style higher education. The Chronicle of Higher Education, 54(3), A1.
Lancaster, H. O., & Seneta, E. (1969). Chi‐Square Distribution. Hoboken, NJ: John Wiley & Sons, Ltd.
Matell, M. S., & Jacoby, J. (1971). Is there an optimal number of alternatives for Likert scale items? Study I: Reliability and validity. Educational and Psychological Measurement, 31(3), 657-674.
Peikes, D. N., Moreno, L., & Orzol, S. M. (2008). Propensity score matching. The American Statistician, 62(3).
Rosenbaum, P. R., & Rubin, D. B. (1985). Constructing a control group using multivariate matched sampling methods that incorporate the propensity score. The American Statistician, 39(1), 33-38.