Analiza podatkov pridobljenih z aplikacijo KAMbi

diplomsko delo

Vuk Stojković (Author), Robert Leskovar (Mentor)

Abstract

Diplomska naloga obravnava podatke, ki so bili pridobljeni s spletno aplikacijo KAMbi.Ta je namenjena dijakom zaključnih letnikov srednjih šol kot pomoč pri izbiri inženirskega študija. Aplikacija uporabnika vodi preko 78 vprašanj o želeni naravi dela in samooceni lastnih kompetenc. Obdelano je bilo 1750 anket. Predstavljena so programska orodja za izvedbo ankete in obdelavo: a) aplikacija KAMbi, ki je napisana z orodjem za malo-kodno programiranje Oracle Application Express in b) jezik R z nekaj najpomembnejših knjižnic, ki so bile uporabljene pri analizi podatkov. Razvit je skript v jeziku R, katerega glavne komponente so: a) branje vhodnih podatkov iz izvoženih tabel baze ali pa direktno branje iz baze podatkov, b) priprava delovnih spremenljivk kot so matrika odgovorov in matrika časov, c) funkcije za izračun dosežka anketiranca po področjih in primerjavo podobnosti odgovorov, d) opisna statistika s histogrami odgovorov ter področij, e) korelacijska analiza ter f) analiza gruč in čiščenje podatkov. Iz opisne statistike je razvidno, da odgovori pogosteje izražajo pozitivne lastnosti anketiranca. Časi reševanja posameznih odgovorov so od 1 do 30 sekund. Pri čiščenju podatkov smo predpostavili, da anketiranec s poprečnimi kognitivnimi sposobnostmi ne more pošteno odgovoriti na vsa vprašanja v manj kot 380 sekundah. Korelacije med odgovori in področji so statistično šibke in nesignifikantne. Analiza gruč je odgovore klasificirala v dve skupini tako v primeru vseh anket kot v primeru očiščenih podatkov.

Keywords

analiza podatkov;KAMbi;R;programiranje;znanost o podatkih;

Data

Language:	Slovenian
Year of publishing:	2021
Typology:	2.11 - Undergraduate Thesis
Organization:	UM FOV - Faculty of Organizational Sciences
Publisher:	[V. Stojković]
UDC:	004.6
COBISS:	83439619
Views:	354
Downloads:	15
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Analysis of data collected by kambi application
Secondary abstract:	The research addresses the analysis of the data obtained with the KAMbi web application. The application is intended as support to choose higher education study programme in engineering. The application guides the user over 78 questions about the desired nature of work and self-assessment of their own competencies. In total 1750 surveys were processed. The software tools for conducting the survey and processing are presented: a) the KAMbi application, which is written in Oracle Application Express, and the R language with most important libraries that were used in the data analysis. A script in R language has been developed. The main components of this script: a) reading input data from exported database tables or reading directly from the database, b) preparation of working variables such as answer matrix and time matrix, c) functions for calculating the respondent’s achievement in areas and comparison of similarity of responses, d) descriptive statistics with histograms of responses and areas, e) correlation analysis, and f) cluster analysis and data cleaning. Descriptive statistics shows that the answers more often express the positive characteristics of the respondent. The times for answering individual answers vary from 1 to 30 seconds. In data cleaning step we assumed that a respondent with average cognitive abilities could not honestly answer all the questions in less than 380 seconds. Correlations between responses and areas are statistically weak and insignificant. The cluster analysis classified the responses into two groups both in the case of all surveys and in the case of cleaned data.
Secondary keywords:	Informacije;Računalniško programiranje;Univerzitetna in visokošolska dela;
Type (COBISS):	Bachelor thesis/paper
Thesis comment:	Univ. v Mariboru, Fak. za organizacijske vede
Pages:	V, 64 f.
ID:	13158438