Lego Challenge Research Report

Blog » Lego Challenge Research Report

Posted on 25 May 2013 02:02

The Lego Challenge

Engl 210.07 (Writing for Engineers)
Section L
Instructor: Andrew Lucchesi
Liudi Yang
Brian Wang
Medwin Chiu
Kari Andresen


A study was conducted on whether or not experience improves collaboration in creating instructions. This was based on Tom Wujec’s study of “The Marshmallow Challenge,” in which he analyzed the characteristics that make up a successful group. For our project, we designed an experiment in which three groups of three members build a LEGO structure based on photographs of multiple views of the completed structure and compose a set of instructions on how to assemble it. They were provided with crayons, papers, pens, pencils, and highlighter markers to create the instructions. The instructions were then given to a designated builder for him/her to build the structure while the group that wrote the instruction watched; the builder was also asked to assess the quality of the instructions. This simulates the experience portion of the study because it helps the groups improve on their instruction writing skills. Then, the groups of instruction writers were required to repeat the instruction-writing process for a different LEGO structure, that would again have to be assembled by the builder. The builder would then compare the quality of the second set of instructions to that of the first. During the experiment, we observed that each member of each group assumed a different role: a person who wrote the instructions, a person who drew the visual aids, and a person who dictated what to write. We also noticed that the time management of the instruction makers improved in the second session as compared to the first session, but their enthusiasm had dropped. We concluded that experience does indeed improve collaborative work overall; however, quality decreases if the process is repeated too much because the task becomes less interesting. Some outside factors that have influenced our results include the different environments in which the groups were tested, the type of relationship between each of the members of the groups, and the extent of knowledge the group members had on the study. This study can provide insight on how to maximize a group’s success by giving them the chance to become experienced and how to provide a  work-conducive environment.


The idea for our study originated from Tom Wujec’s TED Talk on the Marshmallow Challenge, in which groups worked together to build the tallest freestanding structure they can, using uncooked spaghetti, a yard of tape, a yard of string, and a marshmallow to top the structure. While his experiment analyzes what makes one team more successful than another, we wanted to test one of these factors to see if experience improves the success of a group. After some contemplation, we decided to use LEGOs in place of marshmallows, spaghetti, tape, and string because LEGOs are reuseable, cheap, portable, and can easily be put together and taken apart. Because the assignment was to research writing and communication, we designed our experiment so that the test subjects would be organized into groups to compose a set of instructions together on how to build a LEGO structure. Instead of giving them the instructions provided by the LEGO company, we took pictures of the different views of the completed structure for the groups to look at when building it to prevent the groups from simply copying the instructions given by the LEGO company. Because pictures of the completed structure are limited in portraying information about the structure’s interior, we ensured that the structures were simple and that the number of pieces were between 35 and 45.



As stated earlier, our study sought to discover effects of experience performing a specified task on the success of a team when performing the task another time. The task we chose to study consisted of writing and/or drawing instructions for a given LEGO structure for someone else to build.


For our study, we had to gather the following:

  • Three groups of “Instruction Makers”: Each group (Group 1, Group 2, and Group 3) consisted of three people who were familiar with one another. One group was from our Writing for Engineers class and the other two groups were made up of residents of The Towers at CCNY. These groups are the ones writing instructions.
  • Two individual “Instruction Followers”: These two people (Follower 1 and Follower 2) were chosen from our group’s combined social networks. They will be building the LEGO structures using the instructions written by the previous groups.
  • Four different LEGO sets: Sets were chosen to have similar number of pieces and size. Each group was to write instructions for two different sets. See Appendix A for more images. The four sets we chose were:
    • LEGO Space Moon Buggy
    • LEGO City Speed Boat
    • LEGO City Fire Chief Car
    • LEGO Creator Mini Skyflyer
  • Instruction-Making Materials: Groups were provided with the following materials in order to write/draw out their instructions:
    • Several sheets of blank computer paper and looseleaf
    • Colored highlighters
    • Crayons, pens and pencils
  • Digital Camera: Used to take pictures of completed LEGO sets 
  • Computer with webcam: Used to record Instruction Makers while they created instructions and Instruction Followers while they built structures.
  • Food: Consisted of pizza, popcorn, soft drinks, and Buffalo wings. Used as incentive for both parties to fully invest in the experiment.
  • Surveys: Given out after each Instruction Making session.



We performed the experiment three times and each run of the experiment had four phases. Group 1 of the Instruction Makers did their runs in the North Academic Center in an empty classroom while Groups 2 and 3 did their runs in The Towers. Each group performed the experiment at different times. Both Instruction Followers built the structures in The Towers. Follower 1 used the instructions from Groups 1 and 2, and Follower 2 used the set of instructions composed by Group 3.

Phase 1: Instruction Making

Time: 30 minutes

The Instruction Makers were ushered into a room and given a LEGO Set (without the instruction booklet), a MacBook Pro with webcam recording, the Instruction-Making Materials listed above, and food. They were shown pictures of many different views of the completed LEGO structure on the laptop and were given 30 minutes to build the structure as well as create instructions for their corresponding Instruction Follower, using all the materials given to them. We made notes on how group members cooperated. At the end of thirty minutes, the Instruction Makers were given a feedback survey that were meant to be reflective of their process and elicit ideas for improvement.

Phase 2: Instruction Following

Time: 30 minutes

The instructions created in the Phase 1 were given to each group’s corresponding Instruction Follower (listed in “Setup”). For Group 1, this phase occurred several hours after Phase 1 at The Towers rather than in the same classroom. For Group 3, this phase occurred immediately after Phase 1 in the same location. The Instruction Follower was also given the LEGO pieces, food, and up to 30 minutes to assemble the structure using only their corresponding group’s instructions. In order to see how the Instruction Follower followed the directions, the Instruction Makers were given an opportunity to observe the Follower. For Group 1, we took a video of Follower 1 while he attempted to assemble the LEGO. Group 2 was present when Follower 1 attempted to assemble the structure with their instructions. Group 3 was in the same room as Follower 2 while she assembled the pieces from the group’s instructions.

Phase 3: Group Re-evaluation

Time: Variable

The Groups were either given a video of the Follower building the LEGO structure or watched the Follower build the structure in person. While observing the building of the structure, the Groups were allowed to discuss possible changes and improvements in the way they created their instructions. In the case of Group 2 and Group 3, they had to discuss these changes in another room so as to not influence the way their respective Followers built their structures. Each Group was also given their Follower’s completed structure to look at and evaluate.

Phase 4: Repeat

Time: 30 minutes x 2

The Groups repeated Phase 1 with a different LEGO set. The only difference is that the surveys the Group members are required to fill out are slightly different from the first surveys in that they asked for comparisons between the first and second sessions of writing instructions. Then, the second session instructions were passed on to the Instruction Follower, and the Follower again repeats Phase 2. However, the Groups are not required to watch the Followers assemble the structure the second time since their participation in the experiment is no longer required.



The results of our research were highly interesting. We discovered that there were more factors that played in on how well the instructions were written and group work than just experience. Our results are divided into three separate case studies, each one summarizing the results found from our three different groups. These include observations made during the building time, an analysis of the instructions built, and the success of the Instruction Follower following each building period.

Case Study 1

Our first case study involved three classmates who have had previous experience working together from within our class who were tested within a classroom at CCNY. The group was asked to build the moon buggy and the skyflyer. 

During the actual testing process, after the group had finished building the structure, each member assumed different roles while composing the instructions. In this group, one person focused on writing up the instructions, one drew specific pictures to go with these instructions and one aided both of them in their jobs by first dissecting the structure they had completed and then drawing an overall side view of the completed structure. This led to some very interesting and creative instructions, as seen in Appendix B-1. The instructions that this group wrote in the first session (the moon buggy) had one distinctive feature: a combination of colored pictures and written instructions. Each instructional step was accompanied by a corresponding picture. 

When given these instructions, the Instruction Follower had little trouble for the majority of the assembly. At the end, the final structure produced by the Instruction Follower resembled the original structure almost exactly except for the angle of the tail piece. 

The second time through, when asked to build the skyflyer, the group was less enthusiastic about completing the task. However, almost immediately after the time had started, the group began writing the instructions while building the structure. This demonstrated better time allocation within the group. 

The instructions produced this time were less creative (see Appendix B-2). The color-coded pictures were gone and instead replaced by diagrams drawn in pencil. Despite this, the instructions themselves were clearer and the Instruction Follower was able to almost completely replicate the second structure as well as the first time.

Case Study 2

The second case study conducted involved three roommates in the dorms at CCNY. This group was asked to build the chief car and the speed boat in one of the lounges within the dorms. One of the major differences from the other two case studies is the environment in which this group was tested. This group was tested within an environment prone to distractions with people constantly walking in and out of their rooms and other individuals playing on their guitars and having unrelated side conversations.

One of the things we noticed about this group was that the members were extremely relaxed with each other. For the first session the group was asked to build the chief car. Although all three group members realized that they had to complete the task, conversation amongst the group often deviated to unrelated tangents. Despite this, members of the group still took on special roles, similar to the first group. However, the roles differed in that there were 2 people dissecting the structure they had just built and one person doing all the writing. 

This group struggled the most with building the structure and was unable to produce highly viable instructions. In fact, on one of their instructions sheets, they wrote, “Sorry, figure it out yourself.” The instructions that were given however, were effective in conveying the thoughts of the Instruction Makers to the Instruction Follower and he was able to get a general sense of what the overall structure looked but was unable to fully assemble it. 

The second time through, the group was asked to build the speed boat.  During this session, group morale fell drastically and the group dynamic changed. Although the three group members had the experience of completing one of the sessions, their enthusiasm for completing the task had dropped, similar to the first case study. This was particularly noticeable as one of the members spent the entirety of the second session texting on her phone. However, similar to the first case study, the instructions improved and the two functioning group members divided themselves up into the same roles they had previously taken and created the instructions while building the structure. 

Despite the increased efficiency of the group members, they were once again unsuccessful in giving a full set of instructions to the Instruction Follower and he was once again unable to fully recreate the structure. Ultimately, this group was our least effective group.

Case Study 3

The third and final case study conducted involved a group of close friends. These three were tested within an enclosed environment within the dorms, isolated from distractions. The two structures that this group was asked to build were the moon buggy and skyflyer. 

This group was extremely efficient. Right from the start of the first session (creating the moon buggy), the members broke up into separate roles where two people built the structure and recited their process to the third person who wrote detailed instructions with colored pictures. What made this group’s instructions notable was that they highlighted the words that referred to pieces within their instructions so that the Instruction Follower would have a better idea of what pieces are to be utilized when building (see Appendix B-3). This is exactly what happened and the Instruction Follower was able to match the original structure. 

During the second session when the group was asked to build the skyflyer.  the group changed their approach to the task. This changed approach led to a more streamline and effective set of instructions. There were now new roles: one person built the structure, one person wrote out written instructions, and one person drew colored pictures. “Tricky” parts of the structure were drawn out and different views of the completed structure displayed (see Appendix B-4). Once again, the Instruction Follower was able to match the original structure.



From among our case studies, we were able to see common trends between the three groups. 

The first major trend that we saw was the formation of roles within a group. These roles were task-oriented; each member would complete a separate task that were put together to form a complete process of producing instructions. All three groups had an instruction writer, an artist, and someone who dictated what to write and draw. 

The second major trend that we saw amongst the groups is the improvement in time allocation in the second session. Two out of the three groups built their structure first and then proceeded to create instructions in the first session. In the second session, these two groups decided to build and write simultaneously, therefore giving them more time to write the instructions. Because of the additional time, groups were able to write better, less rushed instructions in the second session than the first session, no matter how well or poorly they had performed the first time.

The final trend we noticed was that there was a general loss of enthusiasm in the second session. Groups were very excited to do the experiment in the first session because the entire process was new to them. This led to the output of more creative and more colorful instructions then in the second session, when this enthusiasm had died down. The perfect example of this is seen in the two sessions of Group 1. In the first session, each and every instructional step had a corresponding colored picture. However, in the second set, only a few of the instructions had pictures, and those that did were not colored and were drawn in pencil along with the text.

Other Factors

The data also highly suggests that experience was NOT the only factor that played a role in the success of a group. One of the potential factors that may have affected group success was the relationship between group members. Group 1 was comprised of classmates with prior work history. This resulted in a relaxed but serious work environment and it is reflected in the written instructions produced which were detailed but lighthearted. Group 2 was comprised of three roommates from a dorm building. The three of them had never worked together on a project before and this is reflected in their work ethic and their overall product. The group was dysfunctional at times and the instructions were hastily written. Group 3 comprised of three close friends who understood each other and had probably worked together in the past in a classroom setting. This is reflected in their orderly instructions and the creative liberty shown in both sessions. 

Another factor that may have played a role in success is how groups heard about the experiment. Groups 1 and 3 had members from within the class and therefore would realize how important it would be to be serious about their given task. On the other hand, the members of Group 2 simply volunteered after being given a brief summary of the experiment, and therefore, did not realize the importance of the experiment. 

A third factor that may have played a role in success would be the environment in which the groups worked in. Group 1 and 3 worked in very similar environments. The room was enclosed, they were the only ones in it, and there were minimal outside distractions and noises. These groups could focus on the task at hand: building and writing instructions for their lego structures. On the other hand, Group 2 worked in an environment heavily conducive to distractions. People were walking in and out of the room as it was a public lounge. There were individuals who were simply spending time in the lounge and were not part of the experiment. Finally, food was offered DURING the experiment as opposed to before or after and it therefore became a potential distraction to the Instruction Makers.


This experiment revealed the effects of experience, environment, and relationship on group work. Experience familiarizes the group with the demands of a deadline and it helps them work more efficiently with progressively better time management.  We discovered that group work to accomplish a specific task does improve with more experience. It was consistent with each group we tested that in the second session, time was more efficiently utilized. They gave themselves more time to write the instructions by starting as soon as they started building, recording their progress as they went along, instead of building the structure first and working backward to describe it. They also stayed with their roles because they realized that after the first trial, they were all familiar with their roles and it would take time for another person to get accustomed to that role. None of the groups even discussed the roles; they went straight to doing their specific tasks. 

However, there were also negative effects to this method. Repeating the same tasks for long periods of time counteracts the improvements mentioned above. Exhaustion mixed with the lack of incentive caused a decrease in enthusiasm. At this point, the novelty of the experiment had worn off on the Instruction Makers, especially since they were given the same task to do the second time with the only change being that the structure would be different. In the second session, the drawings were less elaborate and required less effort. This shows how repetitiveness can detract from quality, but not enough to overcome the gains made by learning from experience. A clean, quiet, and private area is the most conducive to work. The environment makes sure that the group is in the right state of mind before completing the task. A professional relationship between workers helps them accomplish the most because it starts them off with a focused attitude and the same goal of accomplishing the task.

The findings of this research project can help companies create teams of workers and provide them with an environment in which they can be the most productive. It is also important for companies to keep in mind that breaks between work times should be provided. The breaks should not be too long so that they lose focus, nor should they be too short so that they lose enthusiasm for good quality. Companies won’t have the problem of providing incentive for work, but these recommendations will better help them assist their own employees. After all, they are paying their employees large sums of money. They would want to benefit as much as they can from the employees.

Future Direction

There are several possible experiment improvements that will yield more consistent data. In order to observe the effects of experience on outcome and group work, all other variables should be standardized. One location should be chosen to conduct the experiment, one that is conducive to work with minimal noise level and distractions. The test subjects should all have the same relationship with their teammates, preferably a professional work relationship. The LEGO sets we chose all had around the same number of pieces but the difficulty level was not the same. Three of the four sets had very distinctive and easy-to-describe pieces whereas the last set had a lot of regular 2x4, 2x2, and 4x1 pieces, which were harder to describe and instruct how to put together. Therefore, we should also take into consideration the age level recommended on the boxes in addition to the number of pieces. Increasing the sample size will also help us make more observations and discern more reliable trends. Also, the groups should be given longer breaks so that they do not lose enthusiasm; this was something we observed in all the groups.

Further research can include exploring the different work ethics of students of different majors. This can help see if they approach the task at hand with different methods. Different age groups will also be interesting to observe. We will be able to see if more years of education makes a difference in group work ethic and results.

Works Cited

Wujec, T., (2010). Tom Wujec: Build a tower, build a team [Video file]. Retrieved from


Appendix A


Appendix B-1


Appendix B-2


Appendix B-3


Appendix B-4


Leave a comment

Add a New Comment
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License