# Factor Analysis

Topics: Factor analysis, Principal component analysis, Tutorial Pages: 17 (2766 words) Published: April 25, 2013
report

Preliminary analysis

Since the data of all variables are non-metric and ordinal, there is no appropriate preliminary analysis can be performed. The descriptive statistics of the variable (table 1) below show that most respondent’s answers falls between 3 to 4 and close to 4 which denote the perception of neutral to agree. That is, most answers are positive indicating the students are not on average dissatisfy with the teaching of the subject.

Figure 1
| |N |Mean |Std. Deviation | |The tutorials are useful and |434 |4.09 |.922 | |relevant | | | | |The tutor is well prepared |436 |4.11 |.962 | |The tutor is knowledgeable |432 |4.19 |.807 | |The tutor is approachable |432 |4.31 |.825 | |The tutor is readily available for |434 |3.93 |.989 | |consultation with students | | | | |In this tutorial I feel free to ask|433 |4.26 |.780 | |questions | | | | |The tutor encourages students to |435 |3.77 |.916 | |think for themselves | | | | |The tutor promotes class discussion|433 |3.51 |1.059 | |The tutor makes the subject |434 |3.40 |1.169 | |interesting | | | | |The tutors spoken English is easily|434 |4.06 |1.248 | |understood | | | | |The tutors explanations are easy to|435 |3.81 |1.101 | |follow | | | | |The lectures and tutorials are well|437 |3.83 |.935 | |co-ordinated | | | | |Working through problem sheets is |432 |3.84 |.910 | |an effective way to understand the | | | | |subject | | | | |The tutorial discussion covers most|435 |3.90 |.920 | |of the important material | | | | |The computer tutorials are well |418 |3.01 |1.165 | |organised | | | | |Valid N (listwise) |395 | | |

Stage 1: Objective of Factor Analysis

Perceptions of a subject’s tutorials on fifteen attribute t1, t2,…,t15 are examined to understand if the perceptions can be grouped and reduced to a smaller number. The focus of this particular analysis is, by grouping the perceptions, the University will be able to see the underlying structure that summaries the opinion of student toward the teaching of the subject. Furthermore, if the fifteen variables can be represented in a smaller number of composite variables, this will help simplify other multivariate analyses such as multiple regressions. This assumes that some sort of underlying order exists in the data being analysed.

Stage 2 Designing a Factor Analysis

Understanding the structure of the perceptions of variables requires R-type factor analysis and a correlation matrix between variables not respondents. All...