Data Mining - Lab Report Example

The survey aimed at developing information on students’ backgrounds for informing teaching practices. Gender, number of previous data science courses by a student, students’ self-assessed data mining efficiency, future career goals, geo-location, and preference for a one-by-one virtual meeting were the study’s variables. Data issues with the collected data, cleaning, and analysis results are discussed. SPSS software was used for analysis.

Data issues and cleaning

Missing data was the most prevalent issue in the data set (Tan, Steinbach, & Kumar, 2006). All data for one participant (ID+ R_wZTAo2AjoAUTWvf) were missing. In addition, data on the number of science-related courses that a student had taken and data on years of professional experience that a student had prior to the course were missing for some of the participants. In addition, data on the expected salary for the first job had unrealistically low values and required cleaning. Means were used to clean data on a previous number of science courses and professional experience and expected salary while the mode was used to clean ordinal data.

The following table summarizes descriptive statistics of the numeric scale variables.

Table 1: Descriptive statistics

Descriptive Statistics

N

Minimum

Maximum

Mean

Std. Deviation

Skewness

Statistic

Statistic

Statistic

Statistic

Statistic

Statistic

Std. Error

previous data science related courses

23

.00

4.00

2.8226

1.05466

-1.098

.481

previous years of professional experience in dara areas

23

.00

21.00

3.6478

4.17722

3.425

.481

expected first salary after graduation

23

29795.78

145000.00

46605.9435

32496.99730

2.341

.481

Valid N (listwise)

23

The three data sets are skewed (p> 0.05) and this means that the media is the best descriptive statistics. The following table shows the statistics.

Table 2: Median for the numeric variables

Statistics

previous data science related courses

previous years of professional experience in data areas

expected first salary after graduation

N

Valid

23

23

23

Missing

0

0

0

Mean

2.8226

3.6478

46605.9435

Median

3.0000

3.6500

29795.7800

Mode

2.82a

3.65

29795.78

a. Multiple modes exist. The smallest value is shown

A majority of the students, therefore, had undertaken about three science-related courses and had about 3.65 years of professional experience in data areas. The students expected a first salary of about $ 29795.78.

A majority of the students (60.9 percent) were fair in data mining efficiency while only 8.7 were good. Only 21.7 percent had much confidence in becoming data analysts after graduation while 56.5 percent were not sure of their positions. Most of the students lived away from campus with 34.8 percent being within driving distance while 52.2 percent lived far away, though within the United States. Most of the students preferred a one-by-one virtual meeting. The following histograms illustrate the distributions.

Graph 1: Data mining efficiency

Graph 2: Interest in becoming a data analyst after graduation

Graph 3: Distance from campus

Graph 4: Preference for a one-by-one virtual meeting

The following table shows significant correlations, based on results in Appendix A.

Table 3: Significant correlations

Previous data science related courses

Previous years of experience in data

0.448

Previous years of experience in data

Expected first salary

0.494

Efficiency

Interest in data analysis

0.489

Correlation between expected salary and level of efficiency identifiny the role of expected salary on motivating students into the subject.

Summary

Majoforrity of the students have sufficient background knowledge in data mining, having done many related courses. They however lack experience in data mining and report average efficiency. Their level of motivation into the data analysis profession is low, their locations are far from the campus, and they prefer one-by-one virtual meetings. A one-on-one approach to learning that focuses on technology for online study is therefore recommended.