DOI: https://doi.org/10.64010/BFYU8740
Abstract
Analysis and intuition were used to examine critical success factors motivating workplace preparation for graduates of the Bachelor of Science in Accounting program. Outcomes achieved upon degree completion for accounting learners, and the basis of proposed model included: stakeholder satisfaction, credential recognition and integrated learning. The first two factors were readily demonstrated to their public, industry and employers through concise data and evidence, while the third, integration of learning was determined by content analysis of student reports by Year 4 Accounting learners in Taxation class. Triangulated evidence from multiple methods validated accounting student’s readiness for career success upon graduation.
Practical Application
Demonstrating the effectiveness of using data-driven and evidence-based approaches to support a theoretical model, a grounded theory method is placed into action. The author elaborates upon the use of content analysis methods. The stakeholder satisfaction portion aligns with ACBSP Standard 3, while the credential recognition section illustrates current approaches to demonstrate the formal value of qualifications. The methods utilized should prove beneficial for business program representatives seeking to communicate the value of their school’s graduates’ readiness for career success. Validation using data and factual evidence offers a means for schools to report to the public effectively, while easing the pathway toward initial accreditation, reaffirmation, writing a quality assurance report, or seeking novel ways to interpret and report on data collected.
Introduction
The SCIL acronym stands for satisfaction, namely stakeholder satisfaction. Credential refers is the award itself and the degree to which it is recognized by the community. Integrated learning is a combination of content, learning tools mastery and motivation to learn. A data and evidence driven approach is used to demonstrate validation of the proposed theoretical model representing Bachelor of Science in Business, Accounting major, and student preparation for career success. Satisfaction is described by analyzing the data from student and employer feedback. Credential is assessed through ascribed status of the award by national and international bodies, industry and trade. Integrated learning is measured using text mining and content analysis. The last portion of the model is elaborated upon in the analysis section of this report. That evaluation is used to leverage computer aided visual analytical methods after the fact-based analysis. The investigator is responsible to deploy their intuition and identify emergent properties toward creation of a unified grounded theory. The target population of this research in Year 4 undergraduate learners from Taxation class in the Accounting major, at an ACBSP Accredited, Region 8, business school.
Definition of Constructs
Credential recognition. Credential recognition is a social construct (Walker, 2007) that refers the degree of ascribed status granted by an academic award to the holder. Having acceptance by industry and trade, as well as national and international accrediting bodies, particularly if these are highly regarded, lends validity to the credential. As such recognition becomes a means to convince the potential employer that the graduate was adequately prepared for success in the workplace.
Integrated learning. “Making connections…between academic knowledge and practice” (Huber, Hutchings & Gale, 2005, p1). Integrated learning is the acquisition of broad-based knowledge about a topic encompassing not only content and contextual knowledge, but also its implementation and how to put knowledge into practice.
Stakeholder satisfaction. A stakeholder is a person, group or organization that has interest or concern in an organization (Freeman & Moutchnik, 2013) and stakeholder satisfaction is a measurement of the degree to which expectationsof an enterprise’s important stakeholders are being met. “Balanced performance measures that are based on actual data compiled from empirical studies and surveys have the most value for tracking company performance over time and in focusing performance improvement efforts.” (Curtice, 2006, p6).
Problem Statement
Business programs face a dilemma on how to demonstrate to their important stakeholders such as accreditors, employers and government that graduates are leaving the program with the required level of knowledge and tools to successfully contribute to the workplace and society. Through this study, the aim is to evaluate through a theoretical model that graduates of the host institution in accounting are adequately prepared for career success.
Research Questions
Three research questions are used to assess the validity and rigor of the evidence and motivate data interpretation toward assessing the preparation level of accounting graduates from the host institution toward career success.
Q1. What evidence exists that demonstrates validity of the accounting credential earned?
Q2. What is the evidence demonstrating stakeholder satisfaction is met with the key stakeholder groups, accounting graduates and their employers?
Q3. What evidence exists from student reflections that can be validated in support of student preparation for career success in accounting that demonstrates engagement with acquisition of content and development of the necessary skills to implement that knowledge as an outcome?
Literature Review
The literature review is a primary tool in deployment of evidence to answer Research Question 1, dealing with official recognition of the credential. The veracity of the agencies providing recognition is examined. The approach of the literature review toward Research Questions 2 and 3 is to lay the necessary platform upon which supporting evidence about stakeholders, content, and motivation of skills can be evaluated.
The Migration Policy Institute (Rabbin, 2013) in a study sanctioned by the European Union but covering the United States, found the US system to be a highly decentralized federal system with a patchwork of organizations involved in credential recognition. Business degrees were less difficult to re-certify than engineering and medical qualifications. CERIS reporting covering migration to Canada and the migrants’ associated credentials. Data was independently assessed by Kelly, Marcelino, and Mullas, (2014) who are also from Canada. These researchers found that “experimental testing has suggested that significant bias exists in employer attitudes towards foreign credentials/experience and minority ethno-racial or religious identities.” (p.38).
Accreditation of higher education institutions is not corruption free, and a recent report asked about the role of quality assurance (QA) in preventing and identifying corruption in higher education institutions, and if QA systems could be constructed that were resistant to corruption. Martin (2016) argued that corruption can be prevented when quality assurance is geared towards control and accountability to stakeholders, is compulsory and deployed as a primary tool for assuring quality. Accreditation quality models are typically based on standards and the accrediting body is in a position to deploy systems where policies, governance structure and practices are required to be met (Martin, 2016).
An example of the accreditation dilemma can be found in Ghana. Ghana has a National Accreditation Board (NAB) and key challenges were identified by Tsilkata and Dotse (2016). At issue was an accreditation controversy fueled by two groups of stakeholders. One group seeing the academic credentials as career tools toward gaining employment, and the other group being more concerned with prestige, self-actualization and seeking the qualification as a means to legitimize themselves in the public view for an acquired position of honor. Both of these groups have flaunted the NAB’s weak accreditation regime and sought credentials that violate the authority of the NAB or where the school or credential granting body has a questionable relationship with NAB. Authors, Tsilkata and Dotse are indeed of Ghanian origin but based at well regarded universities in the United States, Ohio University and Valdosta State respectively.
The United Arab Emirates in its quest to build a knowledge-based society under the patronage of the Ministry of Higher Education and Scientific Research has for the last ten years required institutions to abide by the Commission for Academic Accreditation (CAA) which in 2011 achieved recognition for following best practices from the International Network for Quality Assurance Agencies in Higher Education (INQAAHE). A federation-wide roll-out of a National Qualifications Framework that provided further benchmarks for desired student outcomes and levels of awards was introduced to the higher education institutions in the UAE (CAA, 2011), with a rank-and-file communication at Higher Colleges of Technology occurring in April 2015.
Association of Chartered Certified Accountants (ACCA) is a global professional body for accounting. ACCA provides education recognition in the accounting field. ACCA serves 181 countries, 188,000 members and 480,000 students (accaglobal.com, 2016). Brink and Smith (2012) found the decision between US-based programmatic accreditors for business, i.e. AACSB, ACBSP and IACBE was primarily driven by financial resources. Schools with the greatest resources were most likely to be accredited by AACSB, with ACBSP which has recently become the largest (ACBSP, 2016) falling in the middle and IACBE requiring the least resources.
Computerized analysis methods are gaining in popularity due to high data volumes. Trochim traces the advent of theory-driven approaches, application to program evaluation and creation of theoretical frameworks aiding conceptualization methodology (Trochim, 2016). Studies today can draw upon techniques borrowed from search engine optimization and search engine marketing. One of the early advocates of analyzing the corpus or body of language was Susan Hunston (2002). “The main argument in favour of using a corpus is that it is a more reliable guide to language use than native speaker intuition is” (p. 20). Hunston nevertheless valued intuition stating: “[Intuition] is an essential tool for extrapolating important generalizations from a mass of specific information in a corpus” (p. 22). The method is therefore a mixture a computer analysis and researcher intuition.
According to Berinato (2016), data visualization has become a must-have skill for all managers. This is because a visual abstraction is often times the only way to process the volume and velocity of data that arrives for processing. Furthermore, decision-making increasingly relies upon the ability to make-sense from and interpret of this voluminous amount of what is also known as big-data. Due to open source programs, the internet and proprietary, yet affordable tools, visualization is becoming widely accessible. Access to tools without the deeper understanding of their application can result in producing charts that are inadequate or ineffective.
Berinato proposed that data managers and decision makers ask two questions. “Is the information conceptual or data-driven? and is the statement about the topic declarative or exploratory?” (Berinato, 2016, Page 1). Berinato’s decision model identifies which of four types of visualization goals will be most effective, namely: “idea illustration, idea generation, visual discovery, or general data visualization” (Berinato, 2016, Page 1).
Anzai, and Matsuzawa (2003) used the content analysis methods of word frequency and co-occurrence networks to explore differences in the mission statements of pre-World War II and post-World War II Japanese universities. Their key finding was a greater focus on research in the pre-war university model compared to the post-conflict iterations. A moderating factor was the evolution of post-war universities through privatization to what is known as National University Corporations and a greater focus on profitability.
Data deployment in accreditation is evident in Baldrige Criteria 4.1a, tracking overall organizational performance; WASC Standard 2, quantifying meeting of educational objectives; NCATE Standard 2, implementation of quality assessment systems; and ACBSP Standard 4, measurement of outcomes assessment and student learning outcomes. For example, for ACBSP, some applications would be tracking student information, mid-point assessments, end of program assessments, program review, along with stakeholder oriented data, such as alumni satisfaction and employer satisfaction. (DePape, Lockard & Laramy, 2007).
The method to conduct data visualization with multidimensional scaling was explained by Buja, Swayne, Littman, Dean, Hofmann, and Chen (2008). Tamura (2011) used word analysis, co-occurrence analysis, and categorization to examine a socio-cultural experience shared on an internet discussion forum. Tamura set word analysis limits and reported high frequency words appearing 40 or more times. Using co-occurrence networks and the Jaccard coefficient, Tamura (2011) was able to intuitively create categories.
The Jaccard coefficient is a word frequency algorithm. It divides the frequency of word intersection by the union of word appearance. For example, if the frequency of word a is 4, and frequency of word b is 3, then the frequency of words a and b is 2. As stated as a formula; 2 / (4 + 3 – 2) = 0.4 (Mori, Matsuo, & Ishizuka 2004, p. 2). Tamura (2011) reported that computer coding allowed a researcher to objectively handle large amounts of data, and qualitative information could be easily represented numerically which served to decontextualize the data. Tamura (2011) considered de-contextualization, to be an advantage via removing human factors from a crucial portion of data analysis.
Minami and Ohura (2015) used KH Coder for content analysis of 35 student survey responses. They focused on 20 key responses to Question 11, “How Student’s Attitude Influences on Learning Achievement? An Analysis of Attitude-Representing Words Appearing in Looking-Back Evaluation Texts“. They explored a potential link between learner’s attitude and achievement performance. The authors characterized their study as Education Data-mining (EDD) missed with Knowledge Discovery Data-mining (KDD). Low performers used language and words similar to high performers yet they were poorly implemented while the middle performance group had identifiable word differences. Their approach analyzing a limited number of student responses drew a parallel to the method deployed to analyze content—reflections in the case report on the SCIL model. Dissimilarities were use of a questionnaire, instead of discussion points and class focus on IT—information retrieval. (Minami & Ohura, 2015).
The Handbook of Business Discourse was a useful reference for motivating content analysis in the context of business studies. Systematic replicable methods were described for working with large data sets of text, apportioning words into categories, motivated by uniformly applied computer coding rules. Intuition as well as objective means can be used to process the output. The approach has in common with other data analysis methods, a means to identify patterns, and trends. Content analysis methods are best applied when triangulated with other methods. (Bargiela-Chiappini, Nickerson, & Plancken, 2008, p. 192). Yu, Jannasch-Pennell, and Di Gangi, (2011) concurred. They stated that a computer aided quantitative approach to text-mining of natural language, was reliable and consistent being compatible with grounded theory, provided the researcher remained neutral as categories emerged from the data. Moreover, they justified content analysis as method because it could be used to validate evidence (Yu et al., 2011). Trochim (2016) noted how widespread and interdisciplinary concept mapping methods had become.
KH-Coder has been accepted in the U.S. Court of Law (Posner 2012; Smith, 2014). Text mining and content analysis has been used in over 600 studies, including studies from the Middle East and North Africa, such the mobile learning study of Pelet, Khan, Papadopoulou, and Bernardin (2014). In the tradition of Vygotsky, text mining and content analysis can be used to scaffold collaboration of proficient students helping weaker ones, through adoption of pedagogic processes of performance improvement to construct socio-cultural theories based on student responses from activities (Flowerdew, 2009). “Ultimately Algorithms are a set of instructions followed by computers to solve problems” (O’Neil, 2016). However, O’Neil challenges big data and says rogue algorithms are often with too small of sample sizes with only a few dozen students (2016).
Freeman is the seminal source for Stakeholder Theory which expanded concerned firms’ interest beyond shareholders as the main stakeholder group recognizing that an institution needed to satisfy other parties as well. Freeman remains active today researching the digitization of value creation and how value can be communicated at the expense or formal corporate and social responsibility (CSR) efforts (Freeman & Moutchnik, 2013). Curtice (2006) advocated for a broad stakeholder analysis such as might be deployed with balanced scorecard rather than a unidimensional approach. Stakeholder management has become a competitive dimension for higher education institutions due to the pressures of globalization forcing them to become more entrepreneurial (Marić, 2013). Marić goes on to say that applying listening and learning methods with stakeholders faces two main challenges; 1) identifying critical areas to assess and 2) the actual implementation is difficult (2013).
While primarily examining the UK system, Chapleo and Simms identified three main factors governing stakeholder analysis in higher education. First there is student recruitment and satisfaction, then financial implications which is important given the diverse nature of funding, number of stakeholders and finally assessing stakeholders in terms of their potential impact on the strategic direction (Chapleo & Simms, 2010). University image such as ranking has a relationship to stakeholder satisfaction as well and is particularly important to Middle Eastern business schools. Image was the most important affective factor while university relationships with students, community and companies was the only cognitive factor. (Azoury, Daou and El Khoury, 2014).
To summarize the literature review, the constructs or factors in the proposed SCIL model are based on stakeholder satisfaction, credential recognition and measurements of integrated learning for students leaving college. There is adequate coverage in the literature from the general perspective. However, a gap exists on the specific application of the degree program and specifically at institution in question. Further study will serve to inform on the validity of model by providing evidence in favor or against the SCIL approach.
Method
To answer research questions effectively it is best to uncover material evidence that demonstrates the validity of the program. For Research Question 1, it was necessary to evaluate secondary data about the validity of the credential itself such its’ acceptance by institutional accreditors of the program and qualification. For Research Question 2, secondary data from the institutional effectiveness and quality assurance records, some of which was found to relate directly to stakeholder satisfaction, was evaluated. It was possible to gain an evaluation of the degree of stakeholder satisfaction through interpretation of the metric scalar data. Research Question 3 content analysis of student reflection was used to gain insight into the students understanding of taxation topics, study skills and motivation to learn. This report delves to greater detail in answering Research Question 3 because of quantitatively-based, yet nebulous output using KH Coder. Subjective and highly interpretive means are used to create meaning from the various outputs, namely analysis of, word frequency, hierarchical cluster, co-occurrence networking, multi-dimensional scaling and construct mapping using self-organizing maps.
Participants and sample size.
Higher Colleges of Technology (HCT) learners take the taxation course during either Semester 7 or 8 during Year 4. The sample size of N = 23 (2015) or 28 (2016) represented the entire population for final year accounting majors at HCT-Fujairah Women’s College. Learners undertook a group project and were required to produce a 750-word reflection on what they learned in BUS 4163 Taxation. The course assessment scheme also consists of a series of three faculty-wide exams for which a CSA or common specification is issued. The researcher deployed a corpus-based approach to ethnographic research that is collaborative and contextual. The research strategy was to collect a body of business discourse, namely student reflections from the Year 4 Accounting, Taxation course, and then perform a corpus-based computer aided linguistic analysis.
The study used a variety of outputs in the analysis. Corpus-based approaches are ideal considering the content has no fixed size or content requirements (Bargiela-Chiappini, Nickerson, & Plancken, 2008, 157-158). During ethnographic data gathering the researcher unobtrusively observed participants without intruding upon or manipulating the informants or phenomena (Bargiela-Chiappini, et al. 2008, p 162). Since the sample size is small and the study is not deployed over the 17 campus system it is technically pilot study, for the content analysis portion.
Preparation of Data
The 23 available reflections from AY 2015 and the 28 available from 2016 Year 4 Taxation were assembled into a single plain text document (.txt). HTML code was used to separate each learners’ reflection. Repetitious words such as course or assignment title, or irrelevant details such as student names, student numbers, and characters incomprehensible to the computer program that is used to read in the data were excised from the document. Before KH Coder was run, the software package, Stanford POS tagger (part-of-speech) was uploaded to provide KH Coder a textual frame of reference on how to proceed with the analysis (Pelet, et al., 2014).
Running the Analysis
The reflections were analyzed using KH Coder and the following coding-based techniques word frequency analysis, hierarchical cluster analysis, co-occurrence network, multidimensional scaling and self-organizing map. KH Coder had a user friendly interface. Parameters were specified such as the number of words to use for generating a co-occurrence network. By choosing a lesser number of words, such as 60 that was used for this study it provided a clearer easy to read graph compared to a graph with 100 words. Self-organizing maps take the longest computer processing time. Times to run each procedure took anywhere from a few minutes to several hours.
Word Frequency Analysis
Academic Year (AY) 2015 was compared to AY 2016 through an analysis of word frequency arranged in the order of high to low. Higher-frequency words or keywords (KWs) reveal general themes from the taxation class final reflections. To limit data overload, KH Coder was set to extract 60 words for each analysis (Anzai & Matsuzawa, 2013). The part of speech tagger enabled KH Coder to render a frequency-based output by order of high to low frequency in Excel. By specifying POS tagger in the options menu, KH Coder rendered a top frequency output of words by the categories of noun, proper noun, adjective, adverb, and verb (Higuchi, 2010).
Hierarchical Cluster Analysis
The cluster method used KH Coder’s [Export Document-Word Matrix] which enabled the finding and analysis of word combinations with similar appearance patterns. Results are displayed in a family tree-like dendrogram, which were color-coded by group and connected by tree-branches. Familial relationships could be discerned visually (Higuchi, 2010).
Co-occurence Network.
The displays of co-occurrence networks used circles of varying size to represent keywords extracted from the reflective statements. Line thickness connecting these circles were demonstrative of the strength of these proximity relationship between associated keywords. The strength of association was represented by the thickness of lines connecting the words while the closer proximity, the more likely the words were to co-occur. The structure motivated a visual examination ascertaining the relationship of each keyword to the others and enabled a broader interpretation of the taxation reflections.
Multi-Dimensional Scaling
The method of Higuchi’s (2010) enabled extracted word relationships to be displayed in scatter plots in one, two or three dimensions. The [Export Document-Word Matrix] was a tool to find word combinations with similar appearance patterns. Kruskal graphing was the default mode, with Classical and Sammon modes available if the document became too crowded with overlapping words. Jaccard, Euclid and Cosine were available as a distance estimator options. Cluster analysis allows up to twelve user-specified clusters to be displayed; each with a different color. A two dimensional bubble plot is created which may assist in interpretation of groups.
Self-organizing Map
The mapping method used the KH Coder produced [Export Document-Word Matrix] coupled with the Euclidean distance calculation. As the map grew more complex, plotting more key terminology, a larger volume of text and a greater number of hexagonal coordinates; it can take over an hour to run on an i7 computer. The map can be color-coded whereas light blue (close proximity to its neighbors on the map) changes to white, which changes to shades of pink, which are most distant from its neighbors (Higuchi, 2010).
Results and Discussion
Stakeholder Satisfaction
Employer, graduate and student satisfaction is determined by compiling survey data collected from the respective stakeholder. Specific categories provided include a) Graduate Satisfaction with Academic Programs (by campus, gender), b) Student Satisfaction with Academic Programs, Accounting, c) Graduate Satisfaction (system-wide including analysis and action plans), d) Satisfaction with Campus Facilities, (system-wide including analysis and action plans), e) Student satisfaction with Student Services, (system-wide including analysis and action plans), f) Employer satisfaction with Business Graduates (system-wide including analysis and action plans and g) Employer satisfaction with Business Graduates (by campus, gender). Much of this stakeholder data was included on the 2016 ACBSP QA report.
Employer satisfaction is available for the Business Division, overall, by campus and by gender. Data was not broken out separately for the accounting program, and is an opportunity for improvement. Gender yielded little difference. Current results show men’s graduates at 94% and women’s graduates at 97%. While employer satisfaction varies by campus, overall satisfaction is 95% and that is the current target for 2015-2016. Such data allows outliers to be identified quite easily. For example, Madinat Zayed Women’s College, a smaller Western Region campus significantly underperformed the HCT System in 2013-2014 with an employer satisfaction rating of 28%. Ruwais Women’s College also from the Western Region, underperformed the system average at 72%. An action plan was devised at the system’s program administration level and an attempt made to close the loop. See Appendix 1 for compiled data on Employer satisfaction.
The most recent student satisfaction data is a survey of Bachelor of Science Accounting current students from 2013-2014 and overall satisfaction stands at 93% down slightly from the previous year. Student satisfaction is reported on a program basis for this particular criteria. Graduate satisfaction data is also available which demonstrated a trend of diminished performance of 96% dipping to 87% for the most recent period. An action plan has been devised which includes faculty input. Student Services and Campus Facilities represent opportunities for improvement and can be explained by high staff turnover due to Emiratization in the first case, and in the second case the targets include estimated positive feedback from a massive system-wide investment in facilities for Ras Al Khaimah, Fujairah, Ruwais and Abu Dhabi campuses at a cost of AED 142 million (nearly $40 million USD).
Credential Recognition
Higher Colleges of Technology has institutional accreditation by the Commission for Academic Accreditation. This is UAE Federal accreditation for the entire system (CAA, 2016). The institution also holds the accredited recognition by the Accreditation Council for Business Schools and Programs, effective since 1998. ACBSP is a programmatic accreditation covering all business degrees at HCT, except the discontinued Diploma in Retailing.
The Baccalaureate of Applied Science Degree program has ACBSP accreditation; Bachelor of Applied Science in Business Administration (Accounting). The Higher Diploma (HD) program has ACBSP accreditation and is equivalent to an Associate Degree program: Higher Diploma (HD) Business and Management (Accounting).
Higher Colleges of Technology programs in general are mandated to achieve alignment with the National Qualification Framework (NQF). Regulation is relatively new, having been signed into law in 2010, by President, His Highness Sheikh Khalifa Bin Zayed Al Nahyan who issued Federal Decree No 1. It is understood that under the regulation that programs such as accounting which are a professional qualification, must seek recognition under the National Qualifications Authority (National Qualifications Authority, Qualifications Framework, 2016).
International professional qualifications are recognized in the UAE (National Qualifications Authority, Qualifications Framework, 2016). The Association of Chartered Certified Accountants (ACCA) recognizes most taught portions of the Bachelor of Science in Business Administration (Accounting) and it is certified as an ACCA Accredited Programme. ACCA, 2016). Exemption certificates are granted for papers F1-F9. The only course not granted an exemption certificate was the BUS 4163 Taxation course. The primary reason exemption was not gained on Taxation was a lack of coverage on the United Kingdom (UK) tax code. ACCA as a globally accepted, but UK-based qualification, requires coverage of tax codes to be shifted from covering the United States system which is currently taught, to that of ACCA and the UK (Personal communication, System Course TL, System Course Team Leader Accounting, August 15, 2016). To close the loop, the degree program is being taught out now using the American tax code, and is being re-developed to comply with the British system.
University ranking is another measure of credential recognition and it relates to a school’s reputation with the public in comparison to its peers. The QS Times University Rankings have been released for 2016 and for the Arab Region, Higher Colleges of Technology was ranked 45, slipping from 36 in 2015. Webometrics ranked HCT at number # 8 for United Arab Emirates, while 4International Colleges & Universities ranked HCT at # 2.
Content Analysis using KH Coder
The following section utilizes five methods intrinsic to KH Coder (KH Coder, 2016).
Word Frequency Analysis
Word use was more varied with the two sections of Accounting learners, years 2015 and 2016, compared to those from the 2015 HR class. The findings were unusual in that the HR class participated in an overseas field trip to Greece, including visits to several Greek companies. The expectation was that the experiential learning of an overseas fieldtrip should have impacted the reflections to a larger degree. See Table 1 on the following page illustrating the disparity.
Word Use
Therewere 133 and 95 totalproper nounsrespectively for the two accounting sections.HR learners only deployed 60 such words.Accounting learners were significantly higher in deployment of nouns, adverbs, adjectives and verbs than the HR section.
Table 1
| Part of Speech | HR 2015 | Tax 2015 | Tax 2016 |
| Nouns | 373 | 603 | 617 |
| Proper nouns | 60 | 133 | 95 |
| Adjectives | 153 | 264 | 251 |
| Adverbs | 33 | 59 | 69 |
| Verbs | 189 | 296 | 288 |
| n = | 25 | 23 | 28 |
Note: See Text Analysis Portal for Research (2015).
Table 2 Word Frequency Taxation 2015
| tax | 754 | work | 60 | include | 37 | ||
| income | 299 | help | 57 | make | 37 | ||
| taxation | 197 | taxpayer | 56 | work | 37 | ||
| deduction | 191 | challenge | 54 | gross | 37 | ||
| C or S corporation | 173 | know | 53 | example | 36 | ||
| individual | 146 | people | 48 | person | 36 | ||
| business | 130 | face | 47 | partnership | 35 | ||
| government | 94 | entity | 47 | service | 35 | ||
| learn | 94 | thing | 46 | filing | 33 | ||
| pay | 90 | understand | 45 | important | 33 | ||
| rate | 79 | liability | 44 | exemption | 32 | ||
| different | 79 | calculate | 43 | personal | 32 | ||
| project | 78 | profit | 41 | standard | 32 | ||
| use | 73 | wealth | 41 | married | 31 | ||
| type | 71 | case | 40 | money | 31 | ||
| course | 70 | credit | 40 | company | 29 | ||
| form | 69 | corporate | 40 | information | 29 | ||
| group | 68 | deduct | 40 | revenue | 29 | ||
| country | 63 | apply | 39 | year | 29 | ||
| expense | 60 | time | 38 | like | 29 |
Note: Keyword analysis strategy from Anzai and Matsuzawa, 2013.
Analysis of Taxation Word Frequency
6 out of 10 top words match-up between 2015 and 2016
70% Similarity index between top 60 words
Unique to 2015: form, expense, report, taxpayer, liability, case, profit, credit, apply, gross, filing, important, exemption, personal, standard, married, year, like
Differences were tax related, i.e. AMT, IRS Publication 17 Individual Tax Guide, Standard Deduction versus Itemized, Foreign Tax Credit
Unique to 2016: mean, new, experience, good, member, lot, class, subject, student, need, study, problem, future, provide, liability, way, point, idea, consumption
Differences were teaching and learning related, i.e. Active, IELTS, Extended Outside the Classroom, Experiential
ANOVA: 2015 vs 2016 Accounting, populations differences not statistically significant.
2015 HR vs Accounting sections, differences in population means is statistically significant.
Table 3 Word Frequency Taxation 2016
| tax | 573 | government | 63 | lot | 35 | ||
| taxation | 201 | different | 62 | difficult | 34 | ||
| learn | 147 | make | 60 | class | 33 | ||
| income | 136 | time | 50 | subject | 32 | ||
| business | 131 | country | 49 | student | 31 | ||
| project | 119 | knowledge | 45 | difficulty | 30 | ||
| C or S Corporation | 103 | work | 45 | need | 30 | ||
| course | 99 | mean | 43 | study | 30 | ||
| type | 97 | example | 42 | problem | 29 | ||
| help (ful) | 96 | challenge | 41 | calculation | 29 | ||
| know | 87 | people | 41 | future | 29 | ||
| deduction | 86 | wealth | 41 | provide | 28 | ||
| understand | 80 | include | 41 | liability | 28 | ||
| information | 78 | rate | 40 | way | 28 | ||
| entity | 74 | new | 40 | service | 27 | ||
| calculate | 73 | individual | 39 | payment | 26 | ||
| group | 66 | experience | 37 | point | 26 | ||
| work | 65 | good | 37 | value | 26 | ||
| face | 63 | use | 36 | idea | 26 | ||
| thing | 63 | member | 36 | consumption | 26 |
Note: Keyword analysis strategy from Anzai and Matsuzawa, 2013.
Statistical analysis using ANOVA
Source of Variation SS df MS F P-value F crit___
Between 2015-16 122.5 1 122.5 0.002672 0.960045 5.317655
Between Acct-HR 28037.03 1 28037.03 0.878366 0.376077 5.317655________
Note: 2015 vs 2016 Accounting fail to reject the null hypothesis and populations mean differences are not statistically significant. Accounting vs HR, reject the null hypothesis and conclude that not all of population means are equal.
Hierarchical Cluster Analysis
- The analysis produces a treed dendrogram
- Two objects are merged at every step, two of which that are least dissimilar.
- It is an agglomerative bottoms-up approach i.e. Ward’s (1963) minimum variance method.
- Tree branches can be separated to isolate construct categories.
- Hierarchical Cluster Analysis is often followed with Multi-dimensional scaling
Each cluster of words was submitted into the Google search engine as search terms, and a collective meaning was determined by the researcher based on relevance and meaning of the search results (Hunston, 2010). Hierarchical cluster analysis was similar but not the same with 2015 and 2016 appearing flipped on the visual display of constructs almost a mirror image while individual taxation was present in 2015 and replaced by understanding US tax code for 2016. Cooperative learning and business taxation were common to both models.
Figure 1. Hierarchical Cluster Analysis if Taxation, 2015 vs. 2016.
Figure 2. Hierarchical Cluster Analysis if Taxation for 2015.
Figure 3. Hierarchical Cluster Analysis if Taxation for 2016.
Centrality Co-occurrence Network
The dark pink color, followed by light pink then followed by darker blue indicated degree of centrality. Lines connecting nodes represented network relationships, with a thicker line representative of a stronger connection. For 2015 there were several strong network ties noted. One well networked central theme with three well networked sub themes. For 2016, connections not as strong as for 2015. One well networked central theme with four sub-themes of which two are well networked. The centrality co-occurrence network is shown in Figures 4 and 5 below. Co-occurrence network are known for being aesthetically pleasing, flexible, intuitive to interpret, and user friendly as the graph of the network is being constructed. Furthermore, the output is as symmetrical as possible (Fruchterman, & Reingold 1991).
Figure 4.Reflecting on Taxation 2015 Final Project: Centrality Co-occurrence Network (Higuchi, 2010). Income, deduction, exemption, tax and taxation are connected and most central. There are three groups of interconnected constructs that are distinctly separated from the main group.
Figure 5.Reflecting on Taxation Final Project: Centrality Co-occurrence Network (Higuchi, 2010). Government, service, tax, type and entity, are connected and most central. There are two distinctly separate secondary groupings of constructs, not connected to the main group of central constructs and two sets of pairs that are also distinctly separated from the main group of and the other secondary groupings as well.
Communities Co-occurrence Network
For 2015 the structure of communailiy depicts distinct communities of characteristics centered around a central theme consisting of; tax, income, pay, individual taxable and rate. For 2016 the structure displays four well networked communities and four that are less so. The central theme is relatively weak compared to 2015 and consists of provide, service, payment, value, consumption and wealth. The communality co-occurrence network is shown in Figures 6 and 7 on the following page.
Figure 6.Reflecting on Taxation Communities Co-Occurrence Network, 2015 (Higuchi, 2010).
Figure 7.Reflecting on Taxation Communities Co-Occurrence Network, 2016 (Higuchi, 2010).
Multi-dimensional Scaling (MDS)
Multi-dimensional Scaling is aclassical content analysis approach. As an exploratory method MDS aids the investigator to identify both underlying attributes as well as dimensions. The aim is to interpret how subjects evaluate a given set of constructs or stimuli (Wickelmaeir, 2003).
- Multi-dimensional scaling (MDS) works well for interval and ratio scaled data or for correspondence analysis of nominal data to obtain mapped observations in space.
- It is a graphical way of finding groupings in the data.
- Multi-dimensional scaling is preferred in some cases because MDS has relaxed assumptions of normality, scale data, equal variances and covariances, and sample size.
- The analysis entails, mainly looking for clusters and dimensions.
Figure 9. Four distinct tax related clusters are well delineated and one of these consists of a pair of agglomerated constructs. Tax related constructs are well delineated from the two cooperative learning clusters. Interpretation based on Higuchi, 2010.
- Self-Organizing Map
- Constructs shown in colors such as pink or close to pink such as orange or red, assert a large difference in vectors of neighboring constructs; they are distant. A pink line, in which in these cases there were none; signifies that a vast gulf divides clusters. Shades closer to blue, which includes colors such as purple and green indicate constructs that are proximally related to neighboring constructs on the map. Neutral colors such as gray or white are more neither distal nor proximal (Higuchi, 2010). These descriptions relate to the map in Figure 10 and 11 on the following two pages. The relationship in time and space between constructs, is noted by information given using arrows and text boxes.
Figure 10. Taxation 2016 Self Organizing Map (Higuchi, 2010).
Answers to the Research Questions
Q1. What validation exists that demonstrates validity of the accounting credential earned?
Research Question Q1 (RQ1) should be answered yes. Due to having institutional, globally accepted programmatic and a globally accepted certification credential, RQ1 appears to have achieved face validity. College ranking by QS Times of #45 is not high for the Arab Region and the ranking has slipped since last year when it was #36. Higher Colleges of Technology has not completed CAA mandated accreditation at the program level, albeit in process and nearing completion, while NQF compliance is currently in its earliest stages. Finally, with nine papers exempted by ACCA, and only taxation remaining to be obtained under a revised UK curriculum, acceptance of credential is largely secured by (Higher Colleges of Technology, 2016).
Q2. What validation exists that demonstrates satisfaction is met with the key stakeholder groups, accounting graduates and their employers?
Stakeholder feedback reviewed for Research Question 2 demonstrated that validation was strongly supported. This was in spite of a few instances only where stakeholder feedback was low or eroding. Given the evaluations by program management including the Faculty Academic Committee, action plans are in place and being implemented; attempts to close the loop are demonstrated as underway, such as with large scale investment in facilities.
Q3. What evidence exists from student reflections that can be validated in support of student preparation for career success in accounting that demonstrates engagement with acquisition of content and development of the necessary skills to implement that knowledge as an outcome?
Word frequency analysis supported learners gaining an understanding of taxation constructs and use of language. The co-occurrence network allowed the identification of single most important or central taxation concepts—and differentiated which topics occurred simultaneously in with related constructs. ANOVA statistical procedures demonstrated 2015 and 2016 word-use could not be claimed as different, while taxation could not be claimed as similar to accounting. The hierarchical cluster analysis, produced construct mappings for 2015 and 2016 that were largely mirror images of each other. Construct groups included individual and business taxation that demonstrated that unified material was covered. Multi-dimensional scaling for years 2015 and 2016 illustrated the existence of a cooperative learning construct, distinct and separate from a range of grouped associations of taxation constructs. The self-organizing maps helped view relationships in three-dimensional space with proximal, neutral and distal construct groups.
Implications
The evidence from the study delivered metrics that provided evidence of student preparation for career success in credential recognition and stakeholder satisfaction. The undergraduate Accounting program included the following course learning outcomes (CLOs) which appear to be largely met according to the content analysis performed.
- Differentiate the principles and practices in various tax systems
- Critique the effectiveness of tax system compliance
- Calculate the taxable income and tax liability of individuals
- Calculate the taxable income of business entities
- Calculate the taxable capital gains of assets
- Recommend a business model predicated on a tax system (Higher Colleges of Technology, 2016).
Figure 11. SCIL Model Stakeholder Satisfaction, Credential Recognition and Integrated Learning.
Limitations
This study has numerous limitations. These limitations are primarily related to Research Question 3 and the content analysis portion. Given the current campus, the sample size is small. There is only two not three data sets and thus not enough data to establish a trend when working with the content analysis. The current semester indicated only nine students enrolled in Taxation, although there may be more students next semester. Deploying the study over subsequent semesters will address this limitation. Deficiencies could be rectified by random sampling and testing across a greater number of campuses. Furthermore, once the current degree program is being taught out, a new program is being instituted that will assess primarily proficiency with the Tax code of the United Kingdom. Content analysis evaluation is limited by time, as being snapshot of a particular group of learners and there is difficulty extending this study’s results to other populations due to the lack of a robust random sampling regime.
For Research Question 1 and credential evaluation, external benchmarking was not considered and this type evaluation provides peer-to-peer validation. In addition, several types of data were presented in support of Research Question 2, showed an erosion of stakeholder satisfaction. While actions plans have been implemented in revising the business curriculum and investments made in student services training and facilities, it is important to revisit the areas identified, as opportunities for improvement and ensure that strides were made in improvement as perceived by stakeholders and the loop closed. This study does not include a new metric deployed by the institution measuring outcomes in graduate’s workplace which asks whether graduates are employed in their field.
Future Research
For the immediate short term, tracking subsequent semesters in Fujairah with the content analysis regime will extend the construction of trend data. However, once the program changes over to the UK tax system it will introduce a dynamic change in content. The content analysis methodology could be used to verify closing the loop, as a key program goal will be to obtain exemption on the taxation paper and complete the ACCA certification (Personal communication, System Course TL, System Course Team Leader Accounting, August 15, 2016). Future research may also be extended system-wide and deployed via stratified, clustered random sampling by campus. Another strategy is research should target and define the constructs identified in this study through processes such as factor analysis which has the potential to validate the model. Later research should investigate the nature of graduates’ employment in way that judges the caliber of employment.
Conclusion
This report advanced an argument that used factual evidence and content analysis to support an emergent grounded theory approach to show students were being prepared for career success. The theme is synthesis of fact and knowledge derived from computer coded content. This study has succeeded to propose stakeholder satisfaction, credential recognition and integrated learning as key success factors influencing student preparation for success in the workplace.
References
- 4International Colleges & Universities (2016). University Web Ranking and Reviews Available from http://www.4icu.org/ae/
- ACCA (2016). ACCA education recognition status. Available from http://www.accaglobal.com/content/dam/ACCA_Global/Learning%20Providers/exempt/education-recognition-status.pdf
- Anzai, S., & Matsuzawa, C. (2013). Missions of the Japanese National University Corporations in the 21st Century: Content Analysis of Mission Statements. Academic Journal of Interdisciplinary Studies, 2(3), 197. DOI: 10.5901/ajis.2013.v2n3p197
- Bargiela-Chiappini, F., Nickerson, C., & Plancken, B. (2008) Business Discourse. Retrieved from http://www.palgraveconnect.com/pc/doifinder/10.1057/9780230627710
- Berinato, S., (2016, June). Visualizations That Really Work. Harvard Business Review. Retrieved September 2, 2016 from https://hbr.org/2016/06/visualizations-that-really-work
- Brink, K. E. & Smith, C. (2012), A Comparison of AACSB, ACBSP, and IACBE Accredited U.S. Business Programs: An Institutional Resource Perspective. Business Education & Accreditation, 4 (2), 1-15, 2012. Available at SSRN: http://ssrn.com/abstract=2144954
- Buja, A., Swayne D. F., Littman M. L., Dean, N., Hofman, H., & Chen, L. (2008). Data Visualization with Multidimensional Scaling. Journal of Computational and Graphical Statistics, 17(2), 444-472. doi=10.1198%2F106186008X318440
- Chapleo, C., & Simms, C. (2010). A framework for stakeholder analysis in higher education: A case study of the University of Portsmouth. Business School, University of Portsmouth, England. Available from: http://eprints.bournemouth.ac.uk/20955/1/Stakeholder%20of%20Uni%20(2010)%20Perspectives.pdf
- Commission for Academic Accreditation (2011). Standards for Licensure and Accreditation, Ministry of Higher Education and Scientific Research, United Arab Emirates. Available from https://www.caa.ae/caa/images/Standards2011.pdf
- Commission for Academic Accreditation (2016). Active Programs. Ministry of Higher Education and Scientific Research, United Arab Emirates. Available from https://www.caa.ae/caa/DesktopModules/instPrograms.aspx
- Curtice, R., M. (2006). Stakeholder analysis: The key to balanced performance measures. Process Portfolio Management, BP Trends.com Retrieved from http://www.bptrends.com/publicationfiles/04-06-WP-StakeholderAnalysis-Curtice.pdf
- DePape, J., Lockard, N. & Laramy, R. (2007). Using Accreditation Self-Study Results to Better Understand Student from Recruit through Alumnus. The Center for Teaching and Learning, Preparing Facilitators of Learning for a Diverse World. Take the Credit. Retrieved from http://www.cair.org/wp-content/uploads/sites/474/2015/07/Laramy.pdf
- Flowerdew, L. (2009). Applying corpus linguistics to pedagogy: A critical evaluation*
- International Journal of Corpus Linguistics, 14(3), 393-417. DOI:10.1075/ijcl.14.3.05flo
- Freeman, R. E., Moutchnik, A. (2013). “Stakeholder management and CSR: questions and answers.”. UmweltWirtschaftsForum. 21 (1), 5–9. DOI:10.1007/s00550-013-0266-3
- Higher Colleges of Technology (2016). Retrieved from http://www.hct.ac.ae/
- Huber, M. T., Hutchings, P., & Gale, R. (2005). Integrative Learning for Liberal Education. peerReview, Summer/Fall 7(3/4). Retrieved from https://www.aacu.org/publications-research/periodicals/integrative-learning-liberal-education
- Hunston, S. (2010). Corpora in Applied Linguistics, Cambridge University Press.
- Kelly, P., Marcelino, & L., Mulas, C. (2014). Final report: Foreign credential recognition research synthesis 2009 – 2013. Ceris. Available fromhttp://ceris.ca/wp-content/uploads/2015/01/CERIS-Research-Synthesis-on-Foreign-Credential-Recognition.pdf
- KH Coder, (2016). Open source software, Higuchi, Koichi, Ritsumeikan University, Japan. Available at http://khc.sourceforge.net/.
- Marić, I. (2013). Stakeholder Analisys of Higher Education Institutions. Interdisciplinary Description of Complex Systems 11(2), 217-226. DOI: 10.7906/indecs.11.2.4
- Martin, M. (2016). External quality assurance in higher education: how can it address corruption and other malpractices?, Quality in Higher Education, (22)1, 49-63. DOI: 10.1080/13538322.2016.1144903
- Minami, T., & Ohura, Y. (2015). How Student’s Attitude Influences on Learning Achievement? An Analysis of Attitude Representing Words Appearing in Looking Back Evaluation Texts, International Journal of Database Theory and Application, 8(2), 129-144. Retrieved from ttp://dx.doi.org/10.14257/ijdta.2015.8.2.13
- National Qualifications Authority (2016). International Qualifications. Available from: https://www.nqa.gov.ae/en/QFEmirates/QualificationsFramework/pages/InternationalQualifications.aspx
- National Qualifications Authority (2016). Qualification Framework. Available from: https:// www.nqa.gov.ae/en/QFEmirates/QualificationsFramework/pages/default.aspx#ust seek
- O’Neil, C. (2016). ‘Rogue Algorithms’ and the dark side of big data. Knowledge@Wharton. Wharton, University of Pennsylvania. Available from http://knowledge.wharton.upenn.edu/article/rogue-algorithms-dark-side-big-data/?utm_source=kw_newsletter&utm_medium=email&utm_campaign=2016-09-22
- Pelet, J-E, Khan. J., Papadopoulou, P., & Bernardin, E. (2014). M-Learning: Exploring the Use of Mobile Devices and Social Media, in (Ed) Baporikar, N. (2014). Handbook of Research on Higher Education in the MENA Region: Policy and Practice, 261-296. Hershey, PA: IGI Global.
- QS Times (2016). QS University Ranking Arab Region. Available from: http://www.topuniversities.com/university-rankings/arab-region-university-rankings/2016
- Rabbin, L. (2013). Credential recognition in the United States for foreign professionals. Migration Policy Institute, Washington, DC. Retrieved from www.migrationpolicy.org/pubs/UScredentialrecognition.pdf
- Smith, G. (2014). Data and Intuition, The Conglomerate, Retrieved from http://www.theconglomerate.org/corpus-linguistics/
- Tamura, T. (2011). Application of text-mining methodology to sociological analysis of internet text in Japan. Retrieved from http://www.cajs.tsukuba.ac.jp/monograph/articles/ 01_201103/cajs01_201103_077-097.pdf
- Text Analysis Portal for Research (2015). KH Coder, University of Alberta, Canada. Retrieved
- Tsikata, P. Y., and Dotse, A., K. (2016). The accreditation challenges in transnational educational ecology: The Ghanaian experience, an investigative report. Retrieved from http://www.graphic.com.gh/images/pdfs/acreditation_challenges_full_paper.pdf
- Trochim, W. M., (2016). Hindsight is 20/20: Reflections on the Evolution of Concept Mapping. Evaluation and Program Planning. DOI: 10.1016/j.evalprogplan.2016.08.009
- Posner, R., (2012) Opinion, United States of America, Plaintiff-Appellee, v. Deanna L. Costello, Defendant-Appellant, No. 11-291 U.S. Court of Appeals Seventh Circuit Court.
- Walker, A., J. (2007). Constructing problems and solutions: The social construction of foreign credential recognition as a social problem and policy priority in Canada. Master Thesis, McMaster University, Hamilton, Ontario. Available from https://macsphere.mcmaster.ca/bitstream/11375/10646/1/fulltext.pdf
- Webometric (2016). Ranking Web of Universities Available from: http://www.webometrics.info/en/Asia/United%20Arab%20Emirates%20
- Wickelmaier, F., (2003). An Introduction to MDS. Sound Quality Research Unit, Aalborg University, Denmark, Retrieved from: https://homepage.uni-tuebingen.de/florian.wickelmaier/pubs/Wickelmaier2003SQRU.pdf
- Yu, C. H., Jannasch-Pennell, A., & DiGangi, S. (2011). Compatibility between text mining and qualitative research in the perspectives of grounded theory, content analysis, and reliability. The Qualitative Report, 16(3), 730-744. Available from http://nsuworks.nova.edu/tqr/vol16/iss3/6/
Appendix A
Graduate Satisfaction with Academic Programs
Graduate Satisfaction with Academic Programs by Campus: Business
Campus | AY 2012-13 | AY 2013-14 | AY 2014-15 | Target AY 2015-16 | |||
| % | N | % | N | % | N | % | |
| AAMC | 100% | 2 | 75% | 4 | 80% | 5 | 90% |
| AAWC | 96% | 22 | 82% | 9 | 73% | 3 | 90% |
| ADMC | 99% | 14 | 92% | 29 | 83% | 15 | 90% |
| ADWC | 96% | 18 | 81% | 20 | 100% | 5 | 90% |
| KCWC | 60% | 5 | 75% | 8 | 90% | 4 | 90% |
| DMC | 93% | 161 | 93% | 185 | 89% | 39 | 90% |
| DWC | 97% | 340 | 94% | 231 | 90% | 216 | 90% |
| FMC | 100% | 5 | 100% | 1 | 90% | ||
| FWC | 97% | 101 | 92% | 52 | 73% | 21 | 90% |
| MZWC | 100% | 2 | 80% | 4 | 90% | ||
| RKMC | 100% | 1 | 100% | 1 | 90% | ||
| RUMC | 97% | 6 | 90% | ||||
| RUWC | 100% | 2 | 87% | 3 | 90% | ||
| SMC | 100% | 2 | 67% | 3 | 90% | ||
| SWC | 0.7 | 4 | 0.87 | 3 | 90% | ||
| HCT | 96% | 682 | 92% | 544 | 87% | 323 | |
| Men | 94% | 191 | 93% | 219 | 84% | 64 | |
| Women | 96% | 491 | 92% | 325 | 88% | 259 |
Faculty Response
| 1. Analysis of past trends | In all colleges the satisfaction rate is decreasing. This is alarming, and can be linked to the retention rate. Many students may find alternative higher education. |
| 2. Proposed way forward for improvement | We need to improve our program offerings, and develop new programs. |
Action Plans to Achieve Targets
| Activities | Date | Responsible Person /Unit | |
| 1. | Develop new programs that are attractive to students and differentiate HCT from other universities. | End of the year | FAC |
| 2. | Develop multiple exist points that students have been asking for a long time. | End of the year | FAC |
