Mr HJ Visser (Senior Specialist: Information and Strategic Analysis) (Department of Information and Strategic Anslysis) & Prof DH Tustin (Executive Research Director, Bureau of Market Research) 22 September 2009 University of South Africa (UNISA) STUDENT SATISFACTION MODELLING AN IDEAL TOOL TO LOOPHOLE POTENTIAL THREATS AND WEAKNESSES Southern African Association for Institutional Research (SAAIR) – Forum 2009

Prologue Student Satisfaction Surveys Challenges  Stronger customer-oriented philosophy  Ensure student readiness  Develop retention strategies  Vigilant management approach: satisfaction-intention-retention link CHE - 56 % of South African students drop out NCES - 58 % of students are likely to complete qualification 3 million young South Africans between 18 and 24 years of age are neither in employment, education or training 2

Research methodology… 3 Meta-analysis Research results of previous student satisfaction studies were combined with selected Student Information System (SIS) data student data records Descriptive analysis Mean satisfaction index scores Multivariate analysis Correlation analysis (Bivariate) ANOVA-test Examines ‘internal factors’ impacting on students satisfaction across the various pre-defined engagement areas and student biographics Investigates ‘external’ factors that impact on student success across the various pre-defined student biographics

4 Student registration Student support Administrative Academic services/ products

Hypotheses… 5 Successful students are more satisfied with services than unsuccessful students Successful students’ satisfaction levels differ by gender group Successful students’ satisfaction levels differ by population group Successful students’ satisfaction levels differ by age group Successful students’ satisfaction levels differ by guardian qualification level Successful students’ satisfaction levels differ by employment status Successful students’ satisfaction levels differ by geographic location Successful students’ satisfaction levels differ by entrance category Successful students’ satisfaction levels differ by IT skills category

SATISFACTION SCORES OF SUCCESSFUL AND UNSUCCESSFUL STUDENTS 6 Dimension Unsuccessful studentsSuccessful studentsTotal Index Registration 74 Student Support Admin and Professional Academic 70 Average 7170

BIVARIATE CORRELATION ACROSS ENGAGEMENT AREAS 7 DimensionsRegistration Student support Administration and professional AcademicTotal Registration Student support Administration Academic Total How strong do student success and satisfaction correlate across a selection of biographic variables and how significant is such correlation?

BIVARIATE CORRELATION ANALYSIS 8 VariableCorrelation Sig Gender Male Female Population African Coloured Indian (Asian) White Age years years years years

9 EmploymentCorrelationSig Full-time Part-time Unemployed Full-time student Geographic area Towns Metro/city Entrance category Previous ODL Previous residential No previous experience IT literacy Less IT literate More IT literate Overall

STATISTICAL SIGNIFICANT DIFFERENCES IN SATISFACTION RATINGS 10 Biographic variable RegistrationStudent Support Administration servicesAcademicAverage Index Gender Male Female Total Population group African Coloured Indian White Total Age group Total Statistically significant at a 95% level of confidence

11 Biographic variable RegistrationStudent Support Administration servicesAcademicAverage Index Male guardian qualification Degree Diploma/certificate Grade & ABET No qualification Total Male guardian qualification Degree Diploma/certificate Grade & ABET No qualification Total Employment category Employed full-time Employed part-time Unemployed Full-time student Total Statistically significant at a 95% level of confidence

12 RegistrationStudent Support Administration servicesAcademicAverage Geographic area Towns Metro/city Total Entrance category Previously ODL No previous university Previously residential Total IT literacy levels More advanced IT skills Less advanced IT skills Total Student success Unsuccessful students Successful students Total Statistically significant at a 95% level of confidence

Hypotheses… 13 Research dimension Student successAccepted Gender groupAccepted Population groupRejected Age groupRejected Guardian qualification levelRejected Employment statusRejected Geographic locationRejected Entrance categoryRejected IT literacy Rejected

MOST AND LEAST SATISFIED STUDENT GROUPS 14 Biographic variableMost satisfiedLeast satisfied GenderMalesFemales PopulationAfricanAsian (Indian) AgeOlder studentsYounger students Guardian qualificationNo qualificationDegree qualification EmploymentUnemployedPart-time students GeographicTownsMetro/city Entrance categoryNo previous higher education experiencePrevious ODL students

‘External’ factors 15 Work related (ie stress at work, relevancy of study to workplace) Financial and nonfinancial support and guidance Home environment (ie stress from family and household) Social life (ie sports/spiritual events) Study environment (ie access to library, study space and support) Workload and length of study (ie hours required for studies, period of completion of studies)

External factors 16 Biographic variableLeast satisfied Inhibiting factors (Priority ranking) GenderFemales Work stress, Family stress Finance PopulationAsian (Indian)Work stress, Family stress AgeYounger students Finance, Work stress Lack of support/guidance Guardian qualificationDegree qualificationWork stress, Family stress EmploymentPart-time students Finance, Work stress GeographicMetro/cityWork stress, Finance Enrolment statusPart-time younger studentsWork stress, Finance Entrance categoryResidential university student Work stress Lack access to library Family stress

CONCLUSION 17 STUDENT SATISFACTION MODELLING AN IDEAL TOOL TO LOOPHOLE POTENTIAL THREATS AND WEAKNESSES