I think it is clear that the students have the done requisite work in analyzing the data and creating models. They have explored the data and some preliminary cleansing of the data, mostly by removing samples that are incomplete and performing transformations on certain features.They also considered three different models for predicting the score which allows the group to compare each model and choose the best fit.
However, the presentation of the data makes it extremely hard to make sense of the data. The summary statistics table has values that have a lot of digits which is contrary to the motivation for using summary stats--to understand the data at a glance. I would recommend using scientific notation for this table. Furthermore it is unclear how the data cleansing was executed. Specifically, samples with a sale price of 0 were discarded but those with a sale price of 1 were kept? And as a final recommendation, the graphs need better explanation or should be made simpler. For example the error bars used in one of the plots should have a short blurb explaining their use. Not everyone is familiar with error bars. And the tick marks on the plots also have too many digits, consider presenting the data with a larger scale. For example instead of displaying a tick for a value 2000 in units of 1, use a tick of value 2 in units of 1,000.
I think it is clear that the students have the done requisite work in analyzing the data and creating models. They have explored the data and some preliminary cleansing of the data, mostly by removing samples that are incomplete and performing transformations on certain features.They also considered three different models for predicting the score which allows the group to compare each model and choose the best fit.
However, the presentation of the data makes it extremely hard to make sense of the data. The summary statistics table has values that have a lot of digits which is contrary to the motivation for using summary stats--to understand the data at a glance. I would recommend using scientific notation for this table. Furthermore it is unclear how the data cleansing was executed. Specifically, samples with a sale price of 0 were discarded but those with a sale price of 1 were kept? And as a final recommendation, the graphs need better explanation or should be made simpler. For example the error bars used in one of the plots should have a short blurb explaining their use. Not everyone is familiar with error bars. And the tick marks on the plots also have too many digits, consider presenting the data with a larger scale. For example instead of displaying a tick for a value 2000 in units of 1, use a tick of value 2 in units of 1,000.