Learning About Analysis - Qualitative Data

Qualitative data is any type of data that takes a descriptive form and can be observed but not measured. If given an assessment that requires students to describe a play, characterize a scene, discuss the tragic or describe what is funny, students will give answers such as, “the play was long, the scene held my attention, tragedy is sad and clowns are funny.” These descriptive answers or statements can be observed, but not measured, unless one can code them into numbers. The advantage to qualitative data is that it can describe feelings, thoughts, perceptions and other qualities that are important to us as human beings. The disadvantage to qualitative data is that it cannot be easily summarized, counted, averaged and described to others, unless it is coded.

Qualitative Statements

Qualitative data statements describe student thinking, or judgment. This is important in a variety of assessment objectives. For example, if the objective is to find out how much someone liked a play, one can just make a quantitative scale, for example, from 1-5, and have the students choose the number they want. In doing this, they are making a judgment about the play. However, if knowledge about something that is harder to get at is desired because one has an objective to, for example, find out which part of the play was the most exciting and why, then one would turn to asking students to describe that part and explain why it was the most exciting. One can pick up on parts of student judgment that may not have been found in quantitative data gathering.

Analyzing Qualitative Data

When one collects statements, or even short phrases and one-word answers to the assessments, one needs to perform qualitative data analysis. The first type of analysis needed is a read-through of all the statements and a judgment about the similarities. This is often the only analysis that people perform and the conclusions from this cursory look tend to be superficial. But qualitative analysis has developed some very strong tools in order to uncover more meaningful results. For example, statements, and even parts of statements, can be organized by student and compared to other students. A list can be made of the most often mentioned topics and sorted by number of mentions and instead of concluding that there was a variety of student thinking, good qualitative analysis can provide an in-depth look at student thinking organized by type, frequency, impact and relevance to the objectives of the assessment.

Coding Qualitative Data

Even qualitative statements can be coded into a numbers, providing one can construct a coding system for each statement one receives that will indeed fit into the number system invented to code them. Again, this may involve one’s judgment about student judgment, but sometimes it adds a value to the outcomes that helps the reader to understand the nature of these qualitative statements. For example, qualitative data can be coded and even numbered (yes=1, no=2, etc.)

Simple Coding of Qualitative Data—Table 1

Student ID

Yes or No answer

001

Yes

002

No

003

Yes

When performing a coding on this simple kind of data, one can assign a “1” to Yes and a “2” to No. Now organize the data through a cross tab, or cross-tabulation analysis by counting down and across the tables. For example, the data from the table above could be reorganized in order to analyze the Yes/No data of student scores, and show the percentage of Yes/No answers:

Qualitative Data Coded for Yes/No—Table 2

Student Answer Either Yes or No

Defining the relationship of Yes to No

Yes

2 students answered Yes

No

1 student answered No

Percentage or ratio

2:1 (Yes to No)

The outcome of this analysis can be summarized as, “student answers to a particular assessment were running 2 to 1 in answers that stated, “Yes” to the assessment.”

Harder Statements to Code

Another example would be to ask seven arts teachers to comment on the program in which they were teaching, and these would be the resulting seven statements:

I liked this program (Teacher).
Good program offerings and staff, good staff and student relations (Teacher).
The long classes allowed me to cover a lot of material (Dance Teacher).
This program is a great idea and I am committed to seeing it flourish (Actor).
The courses I taught, after a rocky start, seemed successful (Tutor/coach).
Good field-trip opportunities (Teacher).
Good community showcase performances (Teacher).

The cross tabulations of these qualitative statements could be organized in a variety of ways to help analyze the information needed. For a more powerful analysis of teacher opinion, one could create a table that looks like this:

Teacher Praise for the Program—Cross Tabulation—Table 3

Staff Position

Types of Praise for the Program

 

General

Specific

Complex

Teachers

I liked this program (Teacher)

 

Good community showcase performances (Teacher)

Good field-trip opportunities (Teacher)

Good program offerings and staff, good staff and student relations (Teacher)

Others (Actors, Dance Teacher, Tutor/Coach)

This program is a great idea and I am committed to seeing it flourish (Actor).

The long classes allowed me to cover a lot of material (Dance Teacher).

The courses I taught, after a rocky start, seemed successful (Tutor/coach).

The qualitative statements were coded for the type of praise (general, specific, complex) and from whom the statement was given (teachers vs. others). In coding these, the similarity of the type of comment, and the similarity of who they were, suggested the different types of coding categories (general, specific, complex, teachers, others).

Coding Qualitative Data Leads to Analysis

Coding is one of the most powerful ways to track qualitative statements because a group of seven statements can be turned into a useful table for analysis! This is just the beginning of the analytical power that can be harnessed, even for qualitative data, through cross-tabulation of data.

0