Process Reflections on GA2 Process Reflections
Purpose/ Relevance of this Document
While marking the GA2 process reflections I found that there were numerous commonalities, especially among those groups that were finding the rest of the assignment challenging. Since a few of these factors were directly related to the process used by the groups, I thought it would be a good idea to write a summary reflection, so I created this document. I would like to underline, that this document is not meant as a marking guide, but a summary of my experiences while marking the submissions.
Issues
Following is the list of issues that I have found while reviewing the reflections and corresponding work items.
Style and Content
Several submissions showed typical presentation issues. I thought it is worthwhile to point out these issues, as they are typical of subconscious indicators that reviewers of scientific texts use to see if the submission “smells” of trouble.
Sales Pitches
Some groups described their work as very successful at the start of the reflection, but then indicated in the detail section that they had substantial deviations from the plan, and often had meetings moved, cancelled, or falling apart, and deliverables were late. This discrepancy between the announcement of success and the detail report of failure was often accompanied by data on time consumption in free text, rather than exposed as a table or graph. The reader gets the feeling that the data is either patchy or unfavourable, and that the author attempts to hide it. Such a style easily leaves the reader with the impression that the group is trying to conceal or gloss over unfavourable results. The impression of sales pitches was unfortunately often confirmed when I looked at the results of the review in the issues tables.
Passive Voice
Many groups used passive voice to describe their practices. Since passive voice does not require a subject, its use seems a good choice if you want to avoid naming someone. In context of a description of group practices this can be used obfuscate the responsibility for problems with the work result, or to deny owning up to the practices and approach. It is a good idea to use active voice wherever possible. (See practices section below).Unfortunately, the use of passive coincided with poor results and reports of failure in most cases.
Conditional Sentences
Many groups used second conditional sentences to describe their practices. (Phrases like: “The group were to follow the successful process.”) Second conditional does not indicate if something was actually done. It only indicates a possibility. This implies that the author leaves it open what has really happened. In the cases I have observed in the reflection, the conditional was probably used to obfuscate that a practice was not actually carried out. It is a good idea to use active voice wherever possible.
See the following for reference:
http://en.wikipedia.org/wiki/Conditional_sentence#Counterfactual
Frequently, the use of second conditional actually coincided with poor results and reports of failure in the rest of the reflection text.
Boiler Plates
Some groups used boiler-plate definition text in their plan and review. For example, the roles were being defined explicitly, where a reference to a standard would have been sufficient. Or the process was defined in general terms, rather than as specific commitments of the participants. Such text sections gave the impression that they were serving the role of lorem ipsum (page stuffing) and that the document was delivered mostly as a pro-forma item. Detail reviews usually showed that the methodical approach to the review of groups that used is this style was ad-hoc and haphazard, and did not yield good results.
Colloquial Terminology vs. Domain Terminology
Many teams referred to ‘errors’ in their submissions. In the corresponding standards the problematic aspects of SRS are commonly called either ‘issues’ or ‘defects’. This is a typical case of a discrepancy with domain terminology. If the reader is familiar with a domain, writing text in colloquial terms leaves her or him with the impression that the author is a layperson and does not really know what they are talking about. Finding those terms usually led me to believe that the group that used the terms ‘error’ or ‘bug’ had not investigated corresponding standards or other literature. This lack of know-how could lead to poorer results. Unfortunately, I occasionally found this suspicion confirmed when I looked at the review outcomes.
Method
Some issues with the way the groups approached the assignment also appeared frequently. Here a cause-and-effect relationship becomes visible. The outcomes are more due to the way in which work has been approached, then to the talent of the people involved.
N^3 Sorting
I found that there was a strong correlation between the quality of the SRS reviews and the point in time when the group defined how defects were categorised: Many groups apparently spent substantial energy in first finding ‘as many defects as we could’. This resulted in a mass of entries that had no or no agreed-on order of importance. Now every entry, rather than only the most relevant ones, had to be looked at again and compared with every other entry. For three members this was an effort of O(N^3) where N is the number of entries. This reportedly caused meetings to run long and in an ineffective fashion. Groups that did not use priorities to order their search for defects usually found they were strapped for time. Often they blamed the complexity of the SRS for their lack of progress. They also often suggested altering the review process by dropping steps to make it faster.
Just One Long Night
I found there were a few groups that planned ‘big block’ meetings with lengths above one hour. These groups consistently reported bad progress during those meetings. Often they cited group members leaving and returning, or doing other things as the cause for the lack of progress. The real cause of the issue is probably not the lack of discipline, but the length of the standard attention span (45-60 minutes). Any meeting beyond this length has a very limited effect. Planning a meeting of extensive length consequently does not have a good yield.
No Agreement
Many groups reported members being AWOL for meetings. This was often cited as a reason for slow or lacking progress. In their reflections, none of the groups that cited absenteeism as an issue referred to the group agreements that were submitted at the beginning of the semester in their description of the problem or used the group agreement as a means to mediate the problem. Further research I have done reveals that the groups that reported members being absent consistently did not have an effective group agreement. This shows how failures in the preparation of the group process have consequences in the productive process. These groups should probably review if they want to introduce/update their group agreements.
Bin Bucket
Several groups used group collaboration tools like Google docs to collect defects in a distributed fashion. All of these groups reported that this tool support was helpful, despite occasional technical issues. However, without a person in charge of maintaining the online collection, the results were often pieced together and of inhomogeneous quality. In other words, without a person dedicated to editing the online collection, the result was just the sum of its parts, and not an integrated result.
Recommended Practices
I would recommend the following practices to improve the outcomes of future assignments.
Have a pressure-ready plan – With numbers
When you execute an assignment in a group, plan an effective strategy for dealing with risks and issues before embarking on the work. Group assignments are of a size that requires a whole group to work effectively to succeed. Without a strategy of mitigation the assignment will fail, even if all group members are working in parallel full time.
For example, the assignment came out before the text was released. This order of events implied some risks. What if the text would be too short/ too long? What if there would be hundreds of issues in the text? Ultimately, the plan had to depend on the purpose of the assignment, which was to deliver an SRS review that identified the most pressing defects and open issues, and ideally indentified all issues. Many groups were planning for the ideal case, and their plan did not contain cut-offs or contingencies for a situation where pressure mounts and time runs out.
Use Tools to Ensure Minimal Style Quality
Microsoft Word is not the ideal tool to write technical documents in. But the software has a number of features that can assist you to get started with this unfamiliar type of writing. The following is advice on how to configure Word for technical writing.
When you write any technical document in Microsoft Word
1. check both grammar and style. Enable the checks for all available style options, including those that prevent you from using passive. Rephrase your sentences.
2. use the thesaurus and definition tools and check any term or expression you are unsure about. If in doubt, use a simpler expression that has the exact meaning.
3. enable the display of Readability Statistics. These statistics offer a basic measure of the quality of your writing. Your score should be above 60%.
a. This is the definition of Flesch Reading Ease.
b. You may also want to calculate your Gunning fog index.
In addition, the following two items are highest on my list of advice, as their effect is very visible to the reader.
· If you are not a native language speaker of English, ask a native language speaker to proof-read your text before submission. I do this all the time and it has helped me to deal with the nasty idiomatic idiosyncrasies of the English language.
· Do not repeat material from elsewhere if the intended reader will be familiar with it. Use a citation reference instead.
Some Good Outcomes
You may want to obtain a copy of these groups’ plans and reflections.
· Group T2D submitted a very good review with only very minor faults.
· Group T3F submitted a very good reflection and review. It is exemplary in staying the course with a process and succeeding. Its writing style and presentation are pleasing through terseness and adequate word choice.