3

How do you classify NASA TLX, ANOVA, Matched Pair T-Test, Self-Efficacy Scale, and System Usability Scale? Are these types of data analysis considered quantitative or qualitative?

Please answer whether each is quantitative or qualitative.

Below are a few webpages about these forms of data analysis.

I believe that the NASA TLX, Self-Efficacy Scale, and System Usability Scale are qualitative. While ANOVA and the Matched Pair T-Test are both quantitative.

If I am wrong, please let me know what the correct answer is and why.

Tsundoku
  • 176
  • 1
  • 2
  • 10
Susu Metric
  • 153
  • 5
  • 3
    Good question, except that comparing assessment tools (e.g. NASA TLX) to statistical methods (e.g. ANOVA) is comparing apples to oranges. I suppose the real question is whether quantifying subjective data through scales should be considered qualitative or quantitative data. – Steven Jeuris Apr 09 '15 at 14:16
  • Of course it is. Quantitive analysis isn't magic and it's not much harder to misuse than qualitative analysis is, just has a higher barrier to entry. :v – Christian Hummeluhr Apr 10 '15 at 09:24

2 Answers2

6

ANOVA and t-tests are statistical tests for significance and therefore quantitative.

The other mentioned items are scales (adding numbers to a certain choice) and therefore they can be considered as ordinal scales, and hence as quantitative as they are based on numbers.

The NASA one can be administered by using a sliding scale which can be considered to be a continuous scale, allowing for parametric statistical testing. However, if divided in say 10 choices, it becomes an ordinal scale and non-parametric tests have to be used.

AliceD
  • 20,583
  • 8
  • 49
  • 140
  • I'm still wondering whether NASA TLX can be considered a 'continuous scale'. Not arguing that it generally is considered to be ... – Steven Jeuris Apr 09 '15 at 14:14
  • This doesn't really answer the question. You're describing different forms of quantitative scales (and these types are only really used in behavioral science, to my knowledge, they're not mathematically meaningful). The question is asking how to tell the difference between quantitative and qualitative methodology. – Christian Hummeluhr Apr 10 '15 at 08:47
  • @ChristianHummeluhr - to me it looks like she is referring to specific tests, the more as she mentions scales and statistical tests – AliceD Apr 10 '15 at 09:12
  • Well, she is, but she's not asking what kind of quantitative scale they are. All of these are either quantitative or neither. Measures do not have that property. – Christian Hummeluhr Apr 10 '15 at 09:22
  • Measures can be either, I do not get your point I'm afraid. – AliceD Apr 10 '15 at 20:59
  • @AliceD Measures can't be quantitative or qualitative, but they can be analog or digital. You may be thinking of these? I'm not sure what a "quantitative" measure would be like. In the most basic sense, we can shoot photons at something and record the effect, so a measure can be optical (works via light), but we can't shoot the abstract concepts of twos and threes at something. – Christian Hummeluhr Apr 11 '15 at 07:06
6

It's important to distinguish between measures and analyses, because only analyses can be quantitative or qualitative, not measures.

Measures are, essentially, systematic processes by which we acquire our data, and analyses are processes we use to look at the data. As a rule of thumb, the difference is not hard to find and is given in the name: quantitative analysis deals with numbers, and qualitative analysis doesn't. It looks more complex than it is.

ANOVA and t-tests are both forms of statistical linear regression, for example. They 'concentrate' many standardized observations down to a manageable model we can easily deal with, like an extremely effective form of memory chunking. They're quantitative because they have to do with numbers. Analyses can be quantitative or qualitative, and you can easily tell them apart by whether the technique relies on numbers or judgment.

You didn't actually give any examples of qualitative analysis in the question, but the prototypical example of qualitative analysis is the case study. For various reasons, it's sometimes necessary or just useful to conduct a study based on a single case (e.g., the memory patient HM). These studies can draw meaningful and very useful conclusions based on qualitative analyses like comparative studies, double dichotomy, etc., when we would never be able to reach a meaningful level of statistical power (e.g., because only one or very few patients is/are known to exist in the world).

Measures are neither quantitative or qualitative. Analytic techniques are quantitative or qualitative based on whether or not they work through numbers or human judgment. (Sorry about the formatting, I'm posting from a tablet and will clean it up later.)

Christian Hummeluhr
  • 6,972
  • 4
  • 49
  • 117