Use of Google Earth Engine for Teaching Coding and Monitoring of Environmental Change: A Case Study among STEM and Non-STEM Students

Round 1

Reviewer 1 Report

Overall, the studies you presented in this paper are extremely important. It grants opportunities and opening doors to URM and other students who might not have had similar opportunities. Additionally, it highlights the importance of GEE, and other modeling and related software, both for environmental and general information purposes. My only concern in the abstract and the introduction as they are almost exclusive of one another. The abstract makes no mention of URM or gender discrepancy and the introduction starts abruptly with these factors without providing a proper context. Also, there should be a brief mention of GEE in your abstract as it is an important piece, especially as highlighted in your conclusion.

On other thing to consider is the divide between women and men.  I personally found the statsitical variance of women, especially non-STEM to be of interest.  Why do you speculate that their difference were larger than the men?  Is there a recent comparable study you could match it you?  I personally would add something about this in the conclusion, if not elsewhere and also, in general, include more comparison to previous studies.

Overall, this has abundant merit, but it is important to round it out, even it makes the paper slightly more lengthy.

One additional note: Please revise the formatting for the supplemental document to make it streamlined, unless instructed otherwise by the publisher.

Thanks.

Author Response

Dear Reviewer,

Thank you for your valuable input and time dedicated to reviewing this manuscript. We have revised the manuscript according to your comments and have our responses below.

Abstract: We thank you for your advice on the abstract. We have added text on URM, gender, and included GEE. The abstract now reads as follows:

“Computational skills are advantageous for teaching students to investigate environmental change using satellite remote sensing. This focus is especially relevant given the disproportionate underrepresentation of minorities and women in STEM fields. This study quantified the effects of Earth science remote sensing modules in Google Earth Engine on students’ self-efficacy in coding, understanding remote sensing, interest in science, and a career in environmental research in a STEM and non-STEM class. Additionally, the STEM students engaged in a course-based under-graduate research experience (CURE) on water quality. Satellite imagery was used to visualize water quality changes in coastal areas around the world due to the COVID-19 pandemic shut-down. Pre- and post-surveys reveal statistically significant changes in most students’ confidence to apply coding skills to investigate environmental change and understand remote sensing. The in-tervention was not sufficient to lead to significant changes in interest in science or a career in en-vironmental research. There is great benefit in incorporating remote sensing labs to teach envi-ronmental concepts to STEM and non-STEM students and bolstering the confidence of underrepresented minorities and females in STEM.”

We have included other studies in our discussion involving gender and computational education studies, educational remote sensing studies, and CUREs in other sciences (physics and biology). 

“Similarly, other computer science studies found that though females were less exposed to coding than males, females have a similar or even better aptitude for computing than males [31,35].”

“Scaffolding of computational natural science modules leading to independent research has been previously shown to increase self-efficacy in coding and persistence in STEM [33].”

“Previous CUREs with field and laboratory components in the life sciences helped mo-tivate students to consider or confirm their desire to perform fieldwork and laboratory research while contributing to the Prevalence of Antibiotic Resistance in the Environ-ment (PARE) project [16,37]. Notably, a marine biology CURE increased the science identity in Latinx students [38].”

Once again, we deeply appreciate your diligence in evaluating our work and bringing these suggestions to our attention. Your constructive feedback has been invaluable. Please find attached the revised version of the manuscript. We look forward to any further assessment and remain available to address any additional comments or questions you may have.

Reviewer 2 Report

This is an excellent and timely contribution to STEM/non-STEM education.

Use of satellite data, development of practical computation skills and productive project work in remote and hybrid environments are useful to the community.

Embedding independent research in course structure eliminates a barrier to participation for URM groups in particular. The promulgation of the CURE methodology should be encouraged.

The statistical and qualitative research methodologies are well described.

The authors might consider a possible conclusion from the STEM / non-STEM student data reported in section 3 to the effect that although the interventions provided significant skill enhancements for both student groups, the lack of impact in Q4 and Q5 implies that the non-STEM students are not affected career or interest wise to change directions. Interventions in these personal trajectories are needed long before college entry.

A few minor corrections/suggestions before publication:

1. (line 61-62) It is National Academies of Science, Engineering and Medicine (not Mathematics).

2. Use whole numbers not two decimal places in Non-STEM data in Table 1.

3. Rescale the x-axis of both charts in Fig 1 to span the same range of percentage of responses, which adds some additional nuance to the Non-STEM/STEM comparisons. 

 

 

Author Response

Dear Reviewer,

Thank you for your valuable input and time dedicated to reviewing this manuscript. We have revised the manuscript according to your comments and have our responses below.

1. Thank you for helping us correct this mistake.
2. Thank you for your comment. We changed the non-STEM percentages in Table 1 to whole numbers.
3. Thank you for this suggestion. We have rescaled the plots to have the same axis.

Once again, we deeply appreciate your diligence in evaluating our work and bringing these suggestions to our attention. Your constructive feedback has been invaluable. Please find attached the revised version of the manuscript. We look forward to any further assessment and remain available to address any additional comments or questions you may have.

Reviewer 3 Report

Thank you very much for the opportunity to review this work. Please find my feedback below.

It is an interesting article that quantifies the effects of remote sensing modules in Earth science on students’ self-efficacy in coding, understanding remote sensing, interest in science, and a career in environmental research in a STEM and non-STEM class.
The topic is interesting and related to the journal.

Here are some comments and suggestions:

1) According to Turnitin, manuscript plagiarism is 29%. You have to lower this.

2) I couldn’t find the reference (Callejas et al., 2021) In line 157. Wrong reference style too.

3) When you choose to analyze your data using a paired-sample t-test, a critical part of the process involves checking to make sure that the data you want to analyze can be analyzed using this test. The paired-samples t-test has four assumptions that you have to consider e.g. a) there should be no significant outliers in the differences between the two related groups, b) the distribution of the differences in the dependent variable between the two related groups should be approximately normally distributed, etc.
So, please provide evidence about the assumptions of the t-test. A good example of providing evidence about the statist test assumptions is this: “Chatzopoulos, A., Kalogiannakis, M., Papadakis, S., Papoutsidakis, M., "A Novel, Modular Robot for Educational Robotics Developed Using Action Research Evaluated on Technology Acceptance Model", Education Sciences, 2022”.

4) Does the survey’s questions (Table 2) based on previous research? Please explain, and provide evidence.

5) I could not understand the reason for using these pre-survey questions: students’ favorite number, the name of their first best friend, and the name of their first pet. How do these questions facilitate pairing the pre- and post-surveys? Please explain.

6) The presentation of results is purely descriptive in approach with no connection to the literature, which makes me feel that the literature review is not linked to the data collected. Please provide evidence from the literature to strengthen your results.

Comments for author File: Comments.pdf

Author Response

Dear Reviewer,

Thank you for your valuable input and time dedicated to reviewing this manuscript. We have revised the manuscript according to your comments and have our responses below.

1) Thank you for your comment. We have gone through and edited the manuscript.

2)  Thank you for helping us catch this, we’ve changed the reference to the proper format.

3) Thank you for your suggestion. We have tested our data using the Shapiro test for normality and have decided to use the Wilcoxon signed-rank test for our data due to non-normal distributions. We have added this explanation in our Statistical analysis section and have adjusted the p-values accordingly. 

“Each group was tested for normality using the Shapiro-Wilk test. Our data presented non-normal distributions and thus the Wilcoxon signed-rank test was performed in Python through SciPy.Stats library, at a 95% level of significance for each student grouping and Likert statement. ”

4) Thank you for this question. Our questions were inspired by other computational and CURE educational studies that use Likert questions to gauge students’ self-efficacy with the course/project content, interest in what they’re studying, and career interests. We have cited some of these studies.

“These Likert statements were intended to reveal students’ self-efficacy in coding and remote sensing, their perceptions on science, and career interests as done so in other educational studies [31–33]. ”

5) Thank you for your question. We use these questions at the beginning of each pre and post survey to facilitate matching process and remove duplicates. We used Fuzzy Lookup in Excel to pair the data based on these questions. We have previously used this method in Jay et al., 2019 and have added the reference.

6) Thank you for your suggestion. We have added other studies to our discussion concerning gender, CUREs, and education.

“Similarly, other computer science studies found that though females were less exposed to coding than males, females have a similar or even better aptitude for computing than males [31,35].”

“Scaffolding of computational natural science modules leading to independent research has been previously shown to increase self-efficacy in coding and persistence in STEM [33].”

“Previous CUREs with field and laboratory components in the life sciences helped motivate students to consider or confirm their desire to perform fieldwork and laboratory research while contributing to the Prevalence of Antibiotic Resistance in the Environment (PARE) project [16,37]. Notably, a marine biology CURE increased the science identity in Latinx students [38].”

Once again, we deeply appreciate your diligence in evaluating our work and bringing these suggestions to our attention. Your constructive feedback has been invaluable. Please find attached the revised version of the manuscript. We look forward to any further assessment and remain available to address any additional comments or questions you may have.

Reviewer 4 Report

The paper is well written and, in my opinion, quite ready for publication.

Only three suggestions which might be introduced:

1) Why You start introduction with UMR? First half of the introduction totally did not correspond with the aim and topic of the paper. It should be changed.

2) The description of the CURE is not clear for me.

3) I miss comparison with other, similar approaches.

 

Author Response

Dear Reviewer,

Thank you for your valuable input and time dedicated to reviewing this manuscript. We have revised the manuscript according to your comments and have our responses below.

1. Thank you for this question. We began our introduction with the lack of URM and women in STEM as a motivation for this work. In accordance with another reviewer, we added more language surrounding gender and URM in our abstract.

“Computational skills are advantageous for teaching students to investigate environmental change using satellite remote sensing. This focus is especially relevant given the disproportionate un-derrepresentation of minorities and women in STEM fields. This study quantified the effects of Earth science remote sensing modules in Google Earth Engine on students’ self-efficacy in coding, understanding remote sensing, interest in science, and a career in environmental research in a STEM and non-STEM class. Additionally, the STEM students engaged in a course-based under-graduate research experience (CURE) on water quality. Satellite imagery was used to visualize water quality changes in coastal areas around the world due to the COVID-19 pandemic shut-down. Pre- and post-surveys reveal statistically significant changes in most students’ confidence to apply coding skills to investigate environmental change and understand remote sensing. The in-tervention was not sufficient to lead to significant changes in interest in science or a career in en-vironmental research. There is great benefit in incorporating remote sensing labs to teach envi-ronmental concepts to STEM and non-STEM students and bolstering the confidence of un-derrepresented minorities and females in STEM.”

1. Thank you for your comment. We have added more details on the CURE such as how the students picked their locations, the ORCAA tool, and point readers to the supplemental material where we have included our project templates. 

“In the second half of the term, the STEM students were introduced to the CURE portion of the class. The project involved using a water quality remote sensing tool to investigate changes to water quality due to the COVID-19 anthropause (global pause in human activity). Students were assigned to randomized groups of three to five stu-dents and selected a region of interest for their project. The students used a modified version of the Optical Reef and Coastal Area Assessment (ORCAA) tool created by the Belize and Honduras Water Resources NASA DEVELOP team (https://github.com/NASA-DEVELOP/ORCAA). In this tool, Sentinel-2 MSI and Aqua MODIS satellite imagery are used to visualize water quality changes in coastal areas around the world. Our version was adapted to be shorter and allow for students to modify dates and locations within the code editor as opposed to using the original user interface widgets. Sentinel-2 imagery is used to visualize turbidity, color dissolved organic matter, chlorophyll-a, and normalized difference chlorophyll index. Sea sur-face temperature, chlorophyll-a, Kd(490) (a proxy for water clarity), and particulate organic carbon can be assessed with Aqua. The student groups were free to select any place around the world for the CURE. The students had to identify two nearby loca-tions in their study region for comparison: one where they suspect the anthropause will have affected water quality (i.e., port, large city) and a control site. A total of 15 STEM student groups analyzed the following areas: Thailand, Singapore, San Diego, Cali-fornia, USA, Port of Long Beach in California, Australia, Nigeria, Maldives, India, Houston, Texas, USA, Ho Chi Minh City, Vietnam, Hawaii, USA, and Alaska, USA. The students wrote a final report giving background on how COVID-19 shutdowns affected their area of interest in terms of travel and commerce and attempted to explain water quality findings during and after the COVID-19 shutdown using scientific liter-ature and regional news articles. The report also included their maps and time series plots for each location and water quality parameters. The students augmented their conclusions through supplemental spreadsheets of their time series data and images of their maps. Templates for the project deliverables are located in the Supplementary Materials.”

1. Thank you for this suggestion. We have added similar CUREs, remote sensing, and coding educational studies in the discussion to compare results. 

“Similarly, other computer science studies found that though females were less exposed to coding than males, females have a similar or even better aptitude for computing than males [31,35].”

“Scaffolding of computational natural science modules leading to independent research has been previously shown to increase self-efficacy in coding and persistence in STEM [33].”

“Previous CUREs with field and laboratory components in the life sciences helped motivate students to consider or confirm their desire to perform fieldwork and laboratory research while contributing to the Prevalence of Antibiotic Resistance in the Environment (PARE) project [16,37]. Notably, a marine biology CURE increased the science identity in Latinx students [38].”

Once again, we deeply appreciate your diligence in evaluating our work and bringing these suggestions to our attention. Your constructive feedback has been invaluable. Please find attached the revised version of the manuscript. We look forward to any further assessment and remain available to address any additional comments or questions you may have.