We are conducting a study about the COVID-19 vaccine (in general) hesitancy of the citizens in our city. We are to determine if our intervention is effective to our respondents to change their beliefs and minds, deciding to take the vaccine by measuring their Level of Confidence, Complacency, and Convenience.
We have decided to use "Very confident", "Somewhat confident", "Somewhat not confident" and "Not confident at all" to measure their Level of Confidence. We converted these into numbers, 4, 3, 2 and 1 accordingly. The converted data is then calculated by its mean for each respondent (eg. 2.1-pretest and 3.5 post-test).
Lastly, I performed Ttest to find out the effectiveness of our intervention by its statistical significance and by the increase of the overall mean.
T-test is used to statistically compare two groups of continuous data that are normally distributed. A Likert scale does not produce continuous data (i.e., a response of "4" is not twice in weight as a response of "2"). Thus, at the very least, your approach of treating the Likert scale responses as if they are continuous data is problematic. Discuss the analytic procedure with your research adviser (and this should have been performed during the protocol writing phase).
Please read the following commentary if you have time: Article Analyzing and Interpreting Data From Likert-Type Scales
.Furthermore, consider the excellent responses to a similar question in this thread: https://www.researchgate.net/post/What_statistical_analysis_should_I_use_for_Likert_scale_data.
This is an evergreen issue with Likert scales. In fact there are instances where ordinal scales are approximated as continuous scales for analysis purposes. Please read the following. In my opinion it can be done and these articles can be quoted for reference.
1) Johnson, D.R., & Creech, J.C. (1983). Ordinal measures in multiple indicator models: A simulation study of categorization error. American Sociological Review, 48, 398-407.
2) Norman, G. (2010). Likert scales, levels of measurement and the “laws” of statistics. Advances in Health Sciences Education, 15(5), pp. 625-632. Retrieved from: https://link.springer.com/article/10.1007%2Fs10459-010-9222-y#citeas
For a single item with ordinal scoring, the Mann-Whitney test would be recommended, rather than a t-Test (which requires interval level data).
Thanks for the help!
I forgot to add in the description that we have a scale (4= very confident to 1=not confident it all) to our choices in the questionnaire. Which made our analysis Quantitative and It's best to consult the analysis to someone who has a background in statistics.
Thank you
Rain Jean Eyas Babayson, even if your Likert scale is designed as you mentioned, the result is still unlikely to be quantitative (discrete or continuous)... it is ordinal at best.
According to the likert scale type
If its of 3 grades
Sure. may be No
Its not prefered to transform it to numbers
But if its of
5-6 grades or
The hesitancy hAs many questions each has scale .... so it will be better to be in numbers As it be more easier and logic at the studing of relstionships
According to the likert scale type
If its of 3 grades
Sure. may be No
Its not prefered to transform it to numbers
But if its of
5-6 grades or
The hesitancy hAs many questions each has scale .... so it will be better to be in numbers As it be more easier and logic at the studing of relstionships
According to the likert scale type
If its of 3 grades
Sure. may be No
Its not prefered to transform it to numbers
But if its of
5-6 grades or
The hesitancy hAs many questions each has scale .... so it will be better to be in numbers As it be more easier and logic at the studing of relstionships
According to the likert scale type
If its of 3 grades
Sure. may be No
Its not prefered to transform it to numbers
But if its of
5-6 grades or
The hesitancy hAs many questions each has scale .... so it will be better to be in numbers As it be more easier and logic at the studing of relstionships
According to the likert scale type
If its of 3 grades
Sure. may be No
Its not prefered to transform it to numbers
But if its of
5-6 grades or
The hesitancy hAs many questions each has scale .... so it will be better to be in numbers As it be more easier and logic at the studing of relstionships
According to the likert scale type
If its of 3 grades
Sure. may be No
Its not prefered to transform it to numbers
But if its of
5-6 grades or
The hesitancy hAs many questions each has scale .... so it will be better to be in numbers As it be more easier and logic at the studing of relstionships
According to the likert scale type
If its of 3 grades
Sure. may be No
Its not prefered to transform it to numbers
But if its of
5-6 grades or
The hesitancy hAs many questions each has scale .... so it will be better to be in numbers As it be more easier and logic at the studing of relstionships