This study examines the impact of first-year university students’ expectations and experiences
on their academic performance to enable early strategic interventions. The research
is grounded in various theoretical frameworks, including Astin’s theory of student involvement,
Gardner’s transition theory, Tinto’s theory of student departure and Lizzio’s
framework of five senses of success, providing a comprehensive understanding of the transition
of students to university life. The study follows a six-step Cross-Industry Standard
Process for Data Mining (CRISP-DM). It involves data profiling of student demographics,
academic attributes, expectations and experiences from a sample dataset of 2 054 records
at the University of the Western Cape, South Africa.
Key findings reveal differences in academic performance across different demographics,
with financial support significantly affecting outcomes. Factor analysis identified latent
factors such as effective learning, social well-being, academic support, and access to information.
The study found that the student performance models were not sufficiently
robust for accurate predictions, with F1-scores below 60%. In contrast, academic outcome
models, especially the random forest model, showed more promise, with F1-scores above
70%. Recommendations focus on targeted interventions, comprehensive orientation, enhanced academic support, and fostering an environment for social well-being. The study
highlights the need for a multifaceted approach to student support, emphasising regular
monitoring, evaluation and adaptability in interventions to create a supportive academic
environment.
Hierdie studie doen ondersoek na die impak van eerstejaaruniversiteit-studente se verwagtinge
en ervarings op hul akademiese prestasie ten einde tydige strategiese ingrypings
te aktiveer. Die navorsing is op verskeie teoretiese raamwerke gegrond, wat insluit Astin se
teorie van studentebetrokkenheid, Gardner se oorgangsteorie, Tinto se teorie oor studente
wat opskop en Lizzio se suksesraamwerk van vyf gewaarwordinge. Die navorsing bied dus
’n omvattende begrip van studente se oorgang tot universiteitslewe. Die studie volg ’n
ses-stap, kruisindustrie standaard proses vir dataontginning (Cross-Industry Standard
Process for Data Mining [CRISP-DM]). Dit behels die datasamestelling van studentedemografie,
akademiese eienskappe, verwagtinge en ervarings uit ’n steekproefdatastel van
2 054 rekords van studente aan die Universiteit van Wes-Kaapland, Suid-Afrika.
Deurslaggewende bevindinge dui op verskille in akademiese prestasie oor verskillende
demografie¨e heen, met finansi¨ele steun wat die uitkomste beduidend affekteer. Faktoranalise
het latente faktore soos effektiewe leer, maatskaplike welstand, akademiese
ondersteuning en toegang tot inligting ge¨ıdentifiseer. Die studie het bevind dat studenteprestasiemodelle
nie sterk genoeg was vir akkurate voorspellings nie. F1-tellings
was laer as 60%. Daarteenoor was die akademiese-uitkomstemodelle, veral die ewekansige
bosmodel met F1-tellings van ho¨er as 70% meer belowend. Aanbevelings fokus op
gerigte ingrypings, omvattende ori¨entasie, verhoogde akademiese ondersteuning en die
kweek van ’n omgewing wat maatskaplike welstand bevorder. Die studie vestig die aandag
op die behoefte aan ’n veelvlakkige benadering tot studenteondersteuning en lˆe klem op
gereelde monitering, evaluering en plooibaarheid van ingrypings ten einde n ondersteunende
akademiese omgewing daar te stel.
Patlisiso e e tlhatlhoba tshusumetso ya ditsholofelo tsa ngwaga wa ntlha wa dithuto wa
baithuti ba yunibesithi gammogo le maitemogelo a bone a tiragatso ya seakademiki go ka
ba kgontsha go dira ditogamaano go sa le gale. Patlisiso e e ikaegile mo matlhomesong
a thuto a a mmalwa ao a akaretsang kgopolo ya Astin ya go nna le seabe ga baithuti,
kgopolo ya phetogo ya Gardner, kgopolo ya Tinto ya go tsamaya ga baithuti, le letlhomeso
la maikutlo a katlego a le matlhano la Lizzio, mme tsotlhe tse di tla tlamela ka kitso ya
go tlhaloganya ka botlalo phetogo ya baithuti fa ba tsena mo botshelong jwa yunibesithi.
Patlisiso e e latela magato a le marataro a Thulaganyo ya Tekanyetso ya Kgabaganyomadirelo
ya Kepadatha (CRISP-DM). E akaretsa go dira porofaele ya datha ya palo ya
baithuti, dinonofo tsa seakademiki, ditsholofelo le maitemogelo go tswa seteng ya datha
ya tseosekao ya direkoto di le 2 054 tsa kwa Yunibesithing ya Kapa Bophirima, Aforika
Borwa.
Diphitlhelelo tse di botlhokwa di senola dipharologanyo magareng ga tiragatso ya
seakademiki go kgabaganya dipalo tsa baithuti tse di farologanyeng, le tshegetso ya matlole
eo e amang dipoelo segolo. Tlhatlhobo ya dintlha e supile dintlha tse di fitlhegileng
jaaka go ithuta sentle, boitekanelo jwa loago, tshegetso ya seakademiki, le phitlhelelo ya
tshedimosetso. Patlisiso e e fitlhetse gore dikao tsa tiragatso ya baithuti di ne di sa nonofela
ponelopele e e tlhomameng, ka maduo a F1 a a ka tlase ga 60%. Mo pharologanyong,
dikao tsa dipoelo tsa seakademiki, segolojang sekao sa random forest, se supile tshepiso e
nngwe gape, ka maduo a F1 a a kwa godimo ga 70%. Dikatlanegiso di tsepamisa mogopolo
mo ditsereganyong tse di lebilweng, molebo ka botlalo, tshegetso e e tokafaditsweng ya
seakademiki, gammogo le kgodiso ya maemo a boitekanelo jwa loago. Patlisiso e tlhagisa
tlhokego ya mokgwa wa dikarolo di le dintsi tsa tshegetso ya baithuti, go gatelela tekolo ya
nako le nako, go tlhatlhoba le go fetofetoga ga ditsereganyo go tlhola maemo a tshegetso
a seakademiki.
