Trialling the Use of Google Apps Together with Online Marking to Enhance Collaborative Learning and Provide Effective Feedback

Stage 1 of assignment: collaborative work (Group & subgroup collaboration)

Due to the early scheduling of the practical session in the autumn term, it was not possible to have a lecture to prepare the students prior to the session. Therefore, information was disseminated via Moodle, the University’s Learning Management System, and through a PowerPoint presentation given during one of the first practical sessions of the term.

Each group (W, X, Y and Z) went to a different habitat on campus. The habitats used were as follows:

(For more information about these habitats and site numbers, visit the University’s biodiversity web page.)

Each subgroup collected data from two quadrats within their selected habitat and samples of soil, plants and invertebrates were taken for further analysis in the laboratory. The site was also overviewed to assess different plant species, whether monocot or dicot, and allocated a code to be used by the group. Samples and photos were also taken. This information was used during data collection from the quadrats which were thrown randomly, twice per subgroup. Field measurements taken included a number of different types of plant species, percentage cover for each species, and bare ground.

Additionally, point quadrat was used to assess the height of the vegetation at 10 cm intervals across the quadrat and also to determine the number of hits (e.g. number of leaves contacted before the pin hits the ground when following a vertical path). Temperature of air, plant (IR temperature gun) and soil (temperature probe) were also recorded. Soil was assessed in the field for dampness, odour, and colour. A sample was taken from the centre of the quadrat using a trowel to a depth of 10 cm. The vegetation was put in one labelled sample bag and the soil in a second. In the lab, fresh and dry weight measurements were determined for the vegetation and a subsample of the mixed soil core. Soil samples were also analysed for pH using the method described in the soil test kits (Palintest, 1997); texture was determined from feeling wet and dry samples, and from sedimentation profiles and porosity. Plant samples collected at group level were identified and the range and type of invertebrates found in soil samples were recorded.

Instructions were given via Moodle and verbally during a practical session, so that each student could create a personal Google Account; in addition, a handwritten record of their account details was collected during this practical sessions. Students were instructed to navigate to https://drive.google.com/ to access the Google Apps service.

The lecturer created a folder for the module (BS111) containing four group folders (labelled W, X, Y and Z). Within the group folders there was one folder that was shared at the group level and seven subfolders, each shared only with the relevant subgroup. The Google Drive desktop application was used to duplicate a template folder/file structure, as this proved to be much more time efficient than creating all of the folders and files through the web interface (see this support article for more details about the desktop application and how it works).

Each group folder (X to Z) contained two template documents: a Google Docs file called “BS111 Practical 1 Description of Habitat” (see Supplementary materials 1 and a Google Slides file named “BS111 Photo Submission for Group Work (Supplementary materials 2); the group folder was shared with all students in the group, who were granted full editing rights over the content of the folder. Within each group folder there were a further seven separate sub-folders; these folders contained the Google Sheets document for the data collection task, and were only shared with members of the subgroup. The Google Sheets included pre-formatted tables, with some of the column headers already containing units. The documents at the group and subgroup level also contained inline instructions for the collaborative work (Supplementary materials 3 for example).

Students worked collaboratively at the subgroup and group level; each student downloaded a copy of the Slides and Docs file that was then uploaded to the University’s online coursework submission service (FASER). The group Slides and Docs files were exported as PDF files and the data collection Sheets were exported in Excel format (to make further manipulation possible).

Once this point was reached, the lecturer removed the editing rights of the students in Google Drive, so that further editing was not possible, effectively making all of the collaborative documents “read only”. The documents were then uploaded to Moodle and students used these copies for stages 2 and 3 of the assignment.

The lecturer also established a discussion forum site on the BS111 Moodle page to facilitate student discussion of the fieldwork.

Stage 2 of assignment: Peer assignment

A pro forma for peer assignment was issued to the students (see Supplementary materials 4). Each student peer assessed the other group work (including the subgroup data collection sheet). This was worth 10% of the overall assignment mark.

The lecturer assessed two aspects of the student peer assignment process:

Stage 3 of assignment: Independent work

Printed instructions, also available in Moodle, were issued. The final part of the assignment was individual analysis of the four habitats for patterns, correlations, variability and species richness. This was an open-ended task (see Supplementary materials 5), facilitated by a guidance file (Excel document), which was provided on Moodle for download. Students requested further help and support, which was subsequently given.

The individual assignment files were uploaded to FASER by each student. Electronic marking and feedback was returned at an individual level via FASER.

Stage 1 of assignment: Collaborative work (Group & subgroup collaboration)

The students mainly succeeded in setting up Google Accounts, and the lecturer shared the documents at the appropriate level to enable group work in Google Drive. Handwritten records of the student Google Accounts were nearly complete for two of the four groups; the other two groups emailed the missing account details to the lecturer. There was a good level of online collaboration, evidenced by the group-level documents. The group work demonstrated that students were more aware of the campus grounds, and had an increased understanding of ecology through collecting and analysing the biotic and abiotic characteristics of the campus habitats.

Three groups successfully collected all of the expected field data; the fourth group collected half the expected amount, i.e. from single quadrats rather than two. A considerable amount of data was collected within the two three-hour practical sessions. This involved a lot of collaboration in the field, as well as in the lab, to ensure data sampling and recording was accurate.

Students successfully accessed the documents on Google Drive to create a written description of the habitat. All groups also managed to curate photographs of the habitat site and plant species found using Google Slides. Self-directed learning clearly took place during the creation of the group documents, which surpassed the expectations of the lecturer in terms of overall quality and attention to detail. The descriptions and figures for the submissions were of a very high standard, especially considering the students had no formal training on expected undergraduate academic standards.

An example of the type of data created from one group (X) can be seen in Figure 1; it shows the habitat site ‘The Bluebell Wood’ and one of the plants identified, Plantago major (Figure 2). This data demonstrates that students did research the Minimum Undergraduate Standards required, as evidenced by the correct use of binomial names and the use and position of figure legends. Figure 3a provides evidence that the students engaged in the task, researching and identifying plants to species-level. They produced a high quality document with expected standards, all within six weeks of arriving at university. In general, correct standards were applied to the group level documents. The task instructions were purposely brief to enable students to make their own contribution. Figure 1 and Figure 2 show that some students used this opportunity to deepen their learning and understanding; the inclusion of additional notes shows that several groups were engaging in high quality research into their allocated habitat and the species discovered there (Figure 3a).

Figure 1. Example of student generated Google Slide of habitat from Stage 1 of assignment.

Figure 2. Example of student generated Google Slide of plant from Stage 1 of assignment.

Figure 3.

(a) Example of feedback comments on descriptive task of stage 1 of assignment. (b) Lecturer comments added to student descriptive task in Figure 3a.

The Google Sheets data, collected during the subgroup task, highlighted a greater range of student ability and was less successful overall. Raw and calculated data were expected; the submitted data ranged from no data, incomplete data sets, to data sets complying with Minimum Undergraduate Standards.

The approach required students to be responsible for keeping data safe until it was inputted online, and to understand the Minimum Undergraduate Standards for tables. Evidence suggests that many students were unaware of these expectations, as more than one of the data sheets had no data submitted and many others contained data with the units in the body of the table.

The small size of the subgroups meant that there was greater variation in the standard of work. (This is to be expected for normal distribution of data.) This demonstrates that the subgroups were working within the team that they were assigned to. The large number of data sheets generated (21) meant it was difficult to quickly achieve a single, combined dataset in a format that was suitable for the students to access for the second and third parts of the assignment. The deadline for the assignment was adjusted to take account of the time needed to collate and disseminate the data to the cohort.

Interesting group dynamics were evident from monitoring the Google Drive activity. Members of one group engaged in the process very actively, and some members became natural leaders. These leaders then became a little too controlling and removed the other group members’ editing permissions. The lecturer intervened by sending an email out via Moodle to let them know this was not acceptable behaviour, and reinstated the permissions for all group members. Students used the Moodle forum after this intervention. The students also engaged in group discussions using the Google Apps comments feature. This was particularly useful when students were refining their documents; the comments stream acted as a private, student-led forum for discussion.

Stage 2 of assignment: Peer assignment

The second part of the assignment, the peer assignment of the group work, was aimed at making students aware of the marking process and the criteria that the final individual work would be marked against (questions set and standards expected). The process also exposed individuals to the entire dataset collected. By engaging students in peer review of the document and data collected, it was hoped that individuals would become familiar with the habitats they had not visited, and not just rely on the site that they were actively involved in.

The peer assignment required students to look at the eight group documents, descriptions and figures of the four habitats, as well as the 28 data sheets. A high proportion of students fully engaged in the task (68.4%), i.e. they reviewed the minimum standards and used their observations to peer assess the work of the other three groups (see Figure 4 for an example). This enabled them to achieve greater than a 2.1 mark for the task. It was also found that 52% of students obtaining a first class mark for the peer assignment also obtained a first class mark for their independent work.

Figure 4. Example of a good individual student peer assignment task, stage 2 of assignment.

A number of students struggled with the peer assignment process. The assignments in these cases had arbitrary marking patterns and over-inflated scores, e.g. 5/5, and comments were very brief or absent. 5.3% of students obtained lower second class marks for this second stage of the assignment and demonstrated that they had limited knowledge of expected requirements and types of answers.

The last set of students (26%) did not engage in stage 2 of the assignment at all. There was no evidence that these students knew about Minimum Undergraduate Standards and this was reflected in the work they submitted for stage 3, as the maximum mark obtained for this set of students was 55%.

Stage 3 of assignment: Independent work

The independent work was an open-ended task that challenged the students to work within, and then beyond, their current knowledge base, e.g. use the tutorial guidance to find out about how to carry out averages and correlation. Students also needed to write some accompanying text. The initial tasks (e.g. finding averages) should have been accessible to all students. Stage 3 assumed that stages 1 & 2 had been achieved to a high standard. However, the data sheets between subgroups were not consistent—one group had a complete data set missing and other groups had sheets with incomplete data.

The endpoint of the entire assignment was to carry out correlations to establish if there were any discernible patterns in the data from the different habitats. Students found that the data sheets were not in a suitable format to enable correlation, and had to make difficult decisions on what data to include/exclude and how to standardise the results. As a consequence, the students asked for guidance on sorting and analysing the data. Some individuals became stressed, evidenced by emails sent to the lecturer. An emergency tutorial was run in a computer lab to support these students; however some had overcome these issues by that stage, e.g. they had produced correlation tables and figures which they submitted in their individual assignment (see Figure 5 and Figure 6). During the tutorial, the majority of the attendees felt anxious because they had not achieved the ultimate endpoint of the assignment. As a result, the task was modified to include a simpler outcome that they could achieve. Students had put a huge amount of time into the activity; the lecturer feedback indicated that they had done well and listed achievements for figures and descriptions, and highlighted the fact that most had created tables and figures that mainly complied with expected academic standards.

Figure 5. Demonstrating how a capable student sorted the class data for stage 3.

Figure 6. Demonstrating how a capable student sorted the class data to produce correlation figure (stage 3).

Stage 1 of assignment: Collaborative work (Group & subgroup collaboration)

The division of the cohort into four smaller learning groups and seven subgroups provided an effective and friendly environment for practical work, which provided a good situation to meet the challenge of online collaborative work. However, improved management, e.g. rotating the groups around the laboratory-based activities, would increase the efficiency and reduce the pressure on the limited equipment available in the laboratory.

It was evident from the practicals and data collection stage that there was too much data to collect and analyse; this could be simplified in the future, e.g. reduce sample collection to one quadrat per sub group. This would decrease the time pressure for staff and students during the practical sessions and would also simplify the class data set.

A better system is needed for subgroup data collection, as the method used was cumbersome and not as effective as alternative methods. Other options include the use of a single document at the group level; use of electronic voting devices (clickers) to collect data, or the use of an online form to collect class data in a standard format (Google also have a tool to facilitate this). This would help team leaders provide class data in a usable form. These alternative approaches would make it easier (and quicker) for students to collect and clean the data.

Students collaborated within Google Drive and, in some cases, managed to produce high quality documents (Figure 1 to Figure 3). There was a noticeable variation of style between the four groups. This indicates that students worked within their own group and were not using other mechanisms to exchange information with other students outside of their allocated group. Therefore this part of the assignment met its intended learning outcomes.

The group work, together with the use of comments within Google Docs, demonstrates that students engaged in self-directed learning. The open-ended nature of the assignments allowed the students to deepen their learning, which is evident from research notes added to Figure 1 and Figure 2. Students demonstrated emergent behaviour, which included self-discovery of features in Google Apps, e.g. the discussion stream and document permission settings, which suggests student-student learning took place. Therefore, this met the intended outcome of this approach.

During the group work, it emerged that students were using three methods of electronic communication, e.g. discussing and using comments and emails in Google Drive; using the Moodle forum, and also emailing each other via Outlook. The communication tools within Google Drive were most advantageous, as they acted as a private forum for student discussion. The disadvantage of the Moodle forum was that it was accessible to the whole cohort rather than just the group. Also, more importantly, all lecturers teaching first year modules in the School of Biological Science were able to see comments made; this might colour a lecturer’s view of a particular student when setting and marking their future work. Therefore, it would be wise if the lecturer directed the students to a single method (most likely within the Google ecosystem) as this would avoid such problems.

The success of the three-stage assignment required the rapid creation and collection of Google Account details from students. Overall this worked well, but students who didn’t create accounts did not create any evidence of collaboration, hence did not pass stage 1 of the assignment. Reasons for lack of student engagement were varied: some individuals were reluctant to try something new, others distrusted the online services, and a few didn’t want to put their data in the “Google cloud”. Another possible reason for this poor engagement is the timing of the task; it took place early in the term, so students had to contend with many new things, such as some subgroup members changing degree schemes.

In the future, it would be advantageous to use Google Apps for Education or the Microsoft Office 365 suite of tools, as all members of the University have access to both of these services now. (However, Office 365 would only be used if the collaboration features are as good as those found in Google Drive.)

Using Google Apps for Education would give the lecturer greater control, e.g. the ability to establish student accounts for Google Drive before the start of the practical session. Google takes the education arm of its business very seriously, which should reassure students that their data is safe. Showing the students a short video revealing the security measures that Google enforces at its data centres would perhaps ease any concern, e.g. Inside a Google data center. Getting students to look at the Google Drive privacy policies and terms of service would also be helpful.

Using Google Drive/Apps did make it a little difficult to see which students had done what, as the activity updates are presented as one continuous stream of information. Using Google Apps for Education would also help improve this situation.

It would be possible to provide formative feedback for the stage 1 work rather than wait for formal feedback. This environment is good for self-directed learning, experimentation, and for encouraging informal communication (which will help students engage at a deeper level and help with a more comprehensive understanding of the subject).

Stage 2 of assignment: Peer assignment

The peer assignment provided the link between the group and individual work. It also ensured that the students engaged in data that they were not involved in creating. Those students who engaged in the peer assignment activity went on to do a more thorough individual report (third stage of the assignment). This was in stark contrast to the students who had dealt with the peer assignment superficially, who achieved a much lower score.

Stage 2 enabled students to regain entry into the assignment process if they had failed to complete stage 1. By engaging students at this stage of the assignment, it was hoped that they would become familiar with the habitats they had not visited and not just rely on the site they sampled. Students did analyse all of the data sets, so the peer assignment activity met its intended aim.

Stage 3 of assignment: Independent work

For the third stage of the assignment, some students used the tutorial and managed to achieve all of the expected outcomes, i.e. correlation figures and identified plants to species level through their own endeavours (Figure 5 and Figure 6). It is important to let students know that there is no correct answer expected, and that the assignment is designed to force them to work outside of their comfort zone.

The fact that there were three stages to the assignment allowed any student not succeeding at a particular stage to still be successful at a later stage. Another benefit of this staggered approach was that it required a variety of skills for completion, thus enabled a diverse range of learners to succeed at one or more parts of the assignment.

The overall assignment demanded a high level of engagement, so students who succeeded were clearly involved in a deep learning experience. Many students met this challenge, but some became stressed as they had other assignments to carry out in the same time frame; there was a definite cost versus benefit relationship between the stress created and the learning achieved. To increase the balance towards increased learning, rather than increased stress, a simpler data set together with a more directed final assignment, with some scope for an open-ended approach, will be used in the future.

Finally, the collaborative behaviour that students took part in represents the development of important transferable life skills. The students have been encouraged to use Google Apps for their third year project (but this time they took ownership rather than the lecturer).