Some Glaring Mistakes made by Researchers in Education in Statistical Analysis
1. Some Glaring Mistakes Seen in
Statistical Analysis by
Researchers in Education
Madhavi Dharankar
Asst Prof, School of Education,
YCMOU
dharankar.madhavi@yahoo.com
3. Data Analysis: Some
Observations
• Raw data and frequency tables are given
in chapters
• Percentages, graphs, means taken as
stats techniques
• Meaningless innumerable graphs
10. Observations
• Lack of Identifying Points of
– qualitative analysis
– Triangulation
– Taking analysis to higher level - mixing data
• Lack of Reasoning on
– Nature of data
– Choice of appropriate stats technique
– Strengths and limitations of a technique
11. Taking Analysis Further
• ‘Discussion of Results’ missing
• Difference between findings and
conclusions not clear
• Predictions based on analysis
12. Arguments with the Researchers
(and the Guides)
• “I have not done the calculations. Statistician
has done it for me.”
• “He has used SPSS, computers as a
statistical techniques.” – A guide
Does use of software and/ or help of a
statistician mean saying goodbye to
the basic understanding of analysis in
research?
14. Questions Researchers Could Ask
Themselves
• Why am I drawing a graph? How is it taking
the understanding about the number
‘further’?
• What am I achieving through it, which
otherwise cannot be achieved?
• What am I drawing attention to through this
graph?
• What is the nature of data?
• What does it demand?
• Reading between the numbers
15. Measures at Institutional Level
• University
– Rigor underlined
– Evaluation of thesis by subject expert as well as
statistician (analysis expert)
– Presentations completely focused on analysis
– Formal inputs of stats to both students and
guides
• Collaborations
– Periodical workshops for M Phil and Ph D
research scholars with statisticians
– Analysis clinics (Dr Anil Gore)