search

Dr-Eberle-Zentrum / Data-projects-with-R-and-GitHub

6 stars 3 forks source link

Comments for Sarah by Mohamed #245

Open DrMohamedElsherif opened 2 months ago

DrMohamedElsherif commented 2 months ago

A peer-review on Sarah's project Sarah's project focuses on assessing language proficiency in second language learners using Elicited Imitation Tests (EITs). The objective is to explore the relationship between fluency features and EIT scores, particularly how higher fluency might correlate with better EIT performance.

The project's aim is specific and clear. However, the description would benefit from a more detailed explanation of the relevance of automatic speech recognition (ASR) in assessing language fluency as well as explanation of terms to readers unfamiliar with the field. Additionally, the concept of "messy data" should be clarified. It would be helpful to explain what aspects of the data are messy and what kind of preprocessing or tidying is required. This gives the reader the impression that cleaning the data would actaully be the main task , therefore, more details can help understand the extent of data cleaning required.

Nonetheless, I see here several areas of interesting steps for improvements, For example, It would be beneficial to provide a correlation matrix to show how speech rate, articulation rate, and distance_lv relate to each other. Another area of improvement would be to perform group comparisons to analyze differences in fluency features and EIT scores across the five groups. A third possible expansion is to develop a predictive model to estimate EIT scores based on fluency features.

All in all, Sarah's project sounds interesting and promising, and it has potential for plenty of expansions.

sarahloeber commented 1 month ago

Hi @DrMohamedElsherif,

thanks for your feedback! I definitely agree that I should explain some terms a bit better, especially for people who don't have a background in Linguistics. I'll add some explanations in the background description for that. Since my dataset is not that old yet and I'm working on it myself right now, there are still things emerging from time to time. I will specify what kind of cleaning or manipulation still needs to be done, but it should hopefully not be the main task for this! And thank you for the ideas concerning expansion of the project :)

Best, Sarah