Data Collection
Throughout our data collection, we found the noise level, temperature, and surrounding foot traffic in the different locations on campus that serve food. We were trying to collect this data because the goal of our experiment was to find the best location and time of day to study on campus to optimize the student's time. We believe that noise pollution, temperature, and ambient foot traffic are some of the factors that most affect the quality of a study environment. The noise and the foot traffic can be distracting to a student who is trying to focus on a task or absorb information and being surrounded by a comfortable temperature can be very beneficial as well. Through this data collection, our goal was to pinpoint the optimal time and location for students at Messiah University to study.
Methods
Figure 1: A video of the DecibalX app collecting data as well as the temperature probe and person noting foot traffic
Throughout the experiment, each team member used the process shown in Figure 1 to record our data. This process included first entering the selected central location where the data was to be collected. For the Union, the first-floor central table between the stairs and register was selected. We chose this location because it is the most central location and can be considered to be standard for the entire Union. In Lottie we chose the table between the fountain drinks and register, and the table between the meeting room and entrance for the library all due to their central nature. We would then pull up the excel file on the computer, have the DecibalX app open and ready on the phone, as well as the temperature probe set out. Once we had everything ready to go, we would press reset on the temperature probe and were ready to record. Every ten minutes, we recorded the noise levels for a one-minute interval on the DecibalX app, recorded the temperature displayed on the temperature probe, and counted the number of people walking around within about a fifteen-foot radius. Once the one-minute interval was complete, the DecibalX noise data was uploaded to excel and the amount of foot traffic was recorded. We repeated this process three times for each run.
Figure 2: Flow chart of the data collection process for Foot Traffic
Figure 3: Flow chart of the data collection process for Temperature
Figure 4: Flow chart of the data collection process for Noise (dB)
The flowchart is a visual understanding of the factors our team tested as well as the different levels associated with each. The three response variables we tested were the average sound levels (Figure 4), ambient temperatures (Figure 3), as well as foot traffic during data collection (Figure 2). Each of these variables was tested during the weekday and the weekend which was further broken up into morning and afternoon data collection times seen in Figures 2-4. These collection times consisted of three trials each in which the DecibelX app gathered three, one-minute long decibel recordings.
Figure 5: DecibelX app and temperature probe used
Figure 6: Instruments and Excel sheet for foot traffic collection
The temperature probe is seen in Figures 5 & 6 that we used allowed us to find the temperature of the environment in which we were recording. The probe had an antenna that was attached to the rest of the assembly, which would then display the temperature for us to record. In addition to the antennae and the display, there was a reset button, which we pressed when entering a different location (Figure 8). The reset button allowed us to calibrate the probe so that it could give us accurate readings of the current temperature. The DecibelX app acted as a sound recording instrument. Once the play button is tapped, the app will record a noise level reading for every 0.2 seconds. This allowed us to get nearly 300 data points for each one-minute interval. After the one-minute interval was complete, we could then save the data to the app, and export it as a CSV file, which could be easily opened in Excel (Figure 6).
Because we want to deliver evidence about each of the settings that can be used by students at any point in time, our team’s testing design was built upon the assumption that study conditions will stay the same for each combination of settings. For example, we assumed that the weekday we collected data at the Union in the morning, would be typical for all weekday mornings in the Union. With this assumption in mind, we decided to collect the data for all three of our replications right after one another. This means that we collected all of our Weekday, Lottie, Afternoon data in three consecutive 30-minute blocks. Each 30-minute block consists of three “runs” which are one-minute intervals where we recorded the sound level, temperature, and foot traffic during that one minute. We chose to collect our data in this manner due to the time constraints of the project and because it was within our assumption that conditions would be the same for that combination of settings for any instance. Figure 7 is the final table consisting of the date and time for each trial.
Figure 7: Timing of data collection for entire experiment
Figure 8: Labeled instrument outputs used in data collection
Figure 9: Diagram showing an example of foot traffic radius
Throughout the experiment, we had to directly measure the sound level, temperature, and foot traffic. The sound level as previously mentioned was found using the DecibelX app on our smartphones. As seen in Figure 8 the pause and play button allowed for an easy reset and 60-seconds of data collection to be obtained. For temperature, the ‘outdoor’ digital output (Figure 8) was utilized because it correlated to the more accurate thermocouple seen at the end of the cable. The foot traffic was written down in excel during the one-minute sound interval collection. Figure 9 shows the approximate 15-foot radius in which we would count the number of people entering and exiting during data collection. This was used to gauge whether sound peaking was due to close abrupt noises or considered ambient outside of that range. Throughout this experiment, there were no required calculations or conversions required in terms of analyzing the controllable factors and the corresponding response variables.
Total Cost
We found a greater total cost in our experiment than we had originally predicted. Although the cost of employees decreased due to a shorter time spent collecting data than anticipated, there were complications with our temperature probe. This forced us to need to purchase another one shown below in Figure 10, which overall increased our total cost.
Figure 10: Complete cost of data collection process
Data Sample
Figure 11: DOE means plot for Sound Level from collected data
Figure 11 shows the average sound levels for the various settings we tested. The most important factor is location because this line was the greatest variation in sound level (steepest slope). The Union was the loudest setting overall, most likely due to the number of people who socialize there and the noise from the nearby Café. The Library was by far the quietest setting, which intuitively makes sense considering that most people only go there for the quietest study. There is not a significant change between weekday to the weekend (orange line), however, there is a noticeable change in sound level between morning to afternoon (blue line), with the morning being the quieter of the two.
Figure 12: Temperature Comparison in Union from 9 am to 9 pm
Figure 12 shows the results from our traditional x-y experiment where we took the temperature in the Union, every hour from 9 am to 9 pm. This data shows a slight increase in temperature throughout the day with a significant spike at 6 pm. While we do expect this to be the time of day with the highest temperature, we believe that the intensity of this spike is partly due to instrument malfunction. From this data, we are still able to conclude that the Union stays about the same temperature throughout the day with an increase in temperature at around 5 pm and then a slight decrease in temperature after the sun has set.
Data Conclusions
Now that the data has been collected, we can make several conclusions. One conclusion that can be drawn from our data is that on average, the library is the quietest location by far, followed by Lottie and then the Union. Another conclusion that we can draw is that the library is the warmest location, followed by the Union and then Lottie. Finally, we can see that Lottie is the location that sees the most foot traffic, followed by the Union and then Lottie. This information can be used by any student who wishes to find a place to munch and crunch that best fits their needs.
Overall, our data collection went smoothly. Our tools that were used to take the temperature and noise levels worked very well and made acquiring those measurements simple, while also getting accurate readings. Though a very successful data collection, there were a few things that could have gone better. There were a few times that certain locations where we wanted to test did not open at the times that we had scheduled data collection. Also, the first temperature probe that we used was old, and after a few uses it was no longer working the way that we needed it to, so we had to acquire another one. Finally, a minor issue that we experienced was setting an accurate way to measure foot traffic in various locations. We had said from the beginning that anywhere within a 15-foot radius would be considered ambient foot traffic, but that could be a little bit subjective for each group member.
We would recommend for a future team of engineers who wish to perform the same experiment to add more intervals to their experiment because adding more samples can get a group a better representation of the noise levels, temperature, and foot traffic. Also, future engineers could have multiple people collecting data together, with one person handling the temperature readings and noise levels, while the other count's foot traffic. Finally, future engineers should plan their experiments so that they have more time. With more time, we could collect more data and therefore get close to the true means.
Future Work
Our team designed a very robust and comprehensive testing plan and for this reason, we do not need to collect any more data. Our data provides significant enough evidence to reach conclusions regarding each setting we tested. Moving forward, our team will work to analyze our data and provide statistical evidence for our claims.
Team Member Contributions
To work toward the overall success of this project and track the labor cost, our team tracked individual contributions and distributed work equally throughout various tasks (Figure 13). Aaron spent the majority of his time in excel setting up and organizing data to ensure a smooth data collection process. He also was in charge of interpreting this data and compiling results. Noah completed much of the website content including the cost analysis to determine the total cost of the experiment. Ethan was in charge of compiling all of the information and visuals into the website as well as obtaining the instruments. Every team member contributed toward data collection and was involved allowing for a successful project.
Figure 13: Individual contributions and total time spent
Link to Full Data Set
The button located below will take you to the full excel document. This includes multiple sheets containing different information as follows; the testing assignments and times in which each member collected data, the baseline data obtained, a summary sheet, as well as all of the trials and raw data for each factor.