Week 9: Sampling

Optional: Bernard Chapters 6 and 8

21 thoughts on “Week 9: Sampling

  1. Elise Moody-Roberts

    JJ and I wanted to do a focus group, and we were able to get four students to congregate together at one time, so with two from our class (THANK YOU!) – we had a total of six participants to make a successful group.
    We had sent an email out to over 90 people, and then I called, texted, and met in person with others to see if they could join. I even ran around BiHall minutes before we started to try to recruit more participants. I have a deeper appreciation now for how difficult it is to have a focus group come together at once. I understand the most influential ways to have this happen is to have either or all 1) Good relationships with people 2) Content of the focus group be something they care about or feel will benefit them 3) Additional incentive.
    In thinking ahead to the focus group I would like to have with pregnant women who have an opiate addiction, I am very aware that to have this focus group I must first have developed a strong relationship with people in individual interviews. For such a sensitive topic and vulnerable group, trust and an understanding that the research is to work in their best interest is key. Furthermore, I am considering how to compensate these women for their time, i.e perhaps to provide childcare to make it feasible.
    Our focus group regarding students’ opinions on ‘work’ and ‘play’ at Middlebury College went much smoother than I anticipated. I do not think that anyone in the group was particularly close with or even knew the others, so I was relieved to see the ease in how people offered their thoughts and responded to each other. This is, of course, greatly due to the fact that all six people have some connection and similarity through a shared experience – attending Middlebury College; another key aspect to a successful focus group.
    Our session ran about 40 minutes long, and we did not ask all of our questions on our topic guide. It was a great feeling to know that it could have lasted longer, and I wish it had, although due to peoples’ time commitments we ended it there. We started out with a warm-up introducing ourselves and our study, and then had the participants introduce themselves with their name, class year, and major. We then had them go around again and say what they were involved in or did here. We then played a short film about ‘student life’ on the Middlebury website, and then asked questions as to what and what not the participants identified with, what they thought about the portrayal, what was missing in it, etc. We then moved on to ask more personal questions that ‘dug deeper’ and sought out the specific. We asked why they had chosen to come to Middlebury, investigated into their expectations and values, and into how they evaluated themselves and sought meaning in their work or play.
    The most interesting thing that stood out to me was the consensus in the response about the best part of day. Everyone said it was during mealtimes, or described times when they were neither working nor playing. As I am at my space limit I will end here, noting that I think this observation from our focus group is quite significant and merits considerable examination for our overall study.

  2. Jessica Munyon

    The survey method we used with the map was interesting to say the least. I appreciated the absurd lengths we had to go to in order to ensure the ‘randomness’ of our chosen subjects, but on a campus as small as Middlebury and with a goal to reach as many people as the class the class could, it was obvious that selectivity was a luxury we weren’t able to afford ourselves.

    I immediately went out in search of participants at the conclusion of lab on Thursday. The weather was turning cold and rainy outside and I wanted to get all of my 28 sheets of paper filled out before my scheduled babysitting job at 4:30. Having a specific time limit actually motivated me to complete all of my surveys quickly and made the assignment a bit more amusing.

    Far from amusing, however, was explaining the purpose of the survey for the 30th time in a way that would entice people to take time out of their ‘busy’ schedule to fill out a 90 second survey. I met several people who could do no better than say “I have homework, and really can’t spare the time, sorry.” Really MiddKids? Too much homework to break for 90 seconds and fill out my survey instead of screwing around on facebook for 30minutes? Fine. Suit yourself. I’m clearly ENJOYING freezing my MiddButt off to be awkwardly soliciting participation in one of the most desolate locals on this campus (cough *Mods* cough). I hope your test is as rigorous as your study habits…

    In actuality, I didn’t have that many people refuse to take the survey. Most welcomed the opportunity for a break and to talk to a strange blonde with big, pleading blue eyes. The biggest surprise I had in analyzing the data was how many students that I encountered actually don’t receive financial aid. It took me out of my bubble of assumptions to see just how much of this campus is being funded by the community of students that it provides a service to. All weekend it had me thinking how the administration behaves in accordance with this fact, whether or not students actually work off what they’re paying off, and how much of the ‘party’ mentality robs everyone of a well-rested, competent mindset fresh for Monday morning. No one asks the question that the mother of the children I babysit asked me on Thursday night after I had finished with the surveys: Why is it that we’re all so accustomed to a ‘work hard, play hard’ behavior that never really allows us to be restful and rejuvenated? Isn’t that the point of the reward for rigor in work ethic? I don’t know, anymore. It seems on both accounts, working and playing are more akin to two different jobs, and once one is accomplished, we’re all geared up to get through the next activity on the agenda.

  3. Ansally Kuria (Dot)

    On Thursday during lab my partner (Jenna) and I had already generated the random numbers for the floor and the room numbers for the 10 clusters we were to visit and then split up for a while. I was to go to Kelly and Lang and she was to go to Milliken, after which we would link up again. In my head, I was thinking this was going to be easy and I would be done within no time. However, on reaching Lang not only was the random generated room number empty but all the rooms around it too! I went to the floor below it and there too, nobody answered. So I decided to do the Kelly cluster first. The random floor number was 4. It was empty! I walked into the kitchenette and sat down frustrated. I text Jenna; She seemed to be doing well. I explained my situation and asked her her to text me or call me once she was done working in Milliken so that we could move on to the other clusters. Since I lived in Lang, I could try my luck later. But then I sat there wondering how this was possible so I decided to go through all the floors. If someone answered I gave them the survey, if nobody answered I moved to the next room. I managed to get around seven people in both Lang and Kelly by the time Jenna called. (It was now just after the end of the class period. I actually got many people as they were arriving at their doors).
    Our next cluster was Pearsons, this one was easy. The room number that we had random generated happened to be a double and the two occupants were home. The rooms around it were a double and a triple. Jenna and I then finally split up, I went to Painter and she went to Starr. Later, I was supposed to go to Perkins and she would go to Shannon Street. I did find people in Painter although again I did not find them all on one floor.
    I had no idea where Perkins was, so once I looked it up and realized it was a small house, and not want to frustrate myself again, I figured I really needed to find a time where I would be almost certain of finding people around. Unfortunately my Friday was really full and the weekend was really not a good idea. I decided to go there at 10:30 on Monday night. All but a few of the occupants were away so I managed a cluster pretty quick.
    In all the rooms I went to, only one person said no and that was because he had an exam in an hour, which I really did understand. I was sincerely amazed at just how much people were willing to fill. They asked what it was for and once that I replied it was for a class survey, no more questions were asked. One person did ask me however if we were doing the survey due to the alcohol issues on campus. I also noticed that whenever people were in groups, they read out some of the reasons in 10b a lot and laughed among themselves.
    If a person consented to fill the survey, I handed a copy to them and told them I would be back for it. In one room however, I was asked to wait for the survey which at the beginning was uncomfortable but I just looked around the room…I noticed lots of African postcards and some fabric too. So I asked whether she had been to Africa and she had been. In fact she had been to my country on Safari. So we had a little chat about that which was really nice.
    The entering of the data was more time consuming than I thought but having the number and letter codes did help a lot although I think that the system should have been more standard. For example some questions were not given codes when they should have.
    But all in all I really did enjoy this lab!

  4. Emily Galindo

    I originally thought this project would be a lot of fun but I quickly grew to dislike it and understand why professors work at big universities where there grad students can do this kind of work for them.

    While it was easy to get people to complete the survey once I found them, it was really hard to find people. I actually had only one person ask not to take the survey and that was only because she was very, very sick. For the other dorms it was very time consuming to randomly select rooms throughout the dorm as I could easily be going back and forth between the 1st and 5th floors until I finally find one person in their room. Often times I would pass rooms which obviously had people only to return to them later when they would eventually be randomly selected because no one else was around. On average I would say that for every 5 empty rooms I found someone in a room.

    And although the process of finding people was frustrating, It was interesting to see how people reacted to the survey. Very rarely did anyone ask what the survey was for ( I think about 5 times) and not once did anyone ask how the information was being used or if the information given would be confidential (though people admitted to using drugs, skipping classes, not doing work and presumably underage drinking). More often than not, before I could even begin to explain why I was even there, the person had agreed to take the survey with no questions at all. Also interesting was that i got far fewer negative responses than I expected. I thought I would get a lot more “oh, god, why did i get selected?” attitudes but I only got about 3. Interestingly enough, one of these individuals quickly became more helpful once I told him we went to the same high school, making me wonder, how different would my experience have been if I had looked different, gone to a different hs, was of a different gender, etc. Would people have been less receptive?

    On a completely different note, I foresee a problem when trying to analyze our data as we all entered it slightly differently on the spreadsheet. We’ll see!

  5. Grace Yu

    Surveying people using the map method was interesting for a variety of reasons. First of all, the process of numbering and dividing up the map into sections was interesting because we didn’t use a ruler so the lines may have been a little sketchy. Also, there were a lot of number combinations and it probably would have been smarted to not have so many sections. I never realized how much extra space there is on campus that no one ever goes to and how concentrated student life at Midd is. I wonder if it would have been better if we had just divvyed up all the spaces that actually had people frequent them…but then again, that wouldn’t be accounting for all the random study spots and classes in weird locations. But I felt like it was a random sample of location rather than the student population.

    Getting to the locations wasn’t all to random either because I deliberately chose times that I knew people would be there so that I wouldn’t have to wait in the cold. It was really interesting seeing the people who did respond and the people who didn’t. I think this survey from out group probably excludes the people who are super busy and also people who don’t enjoy stranger interaction because some of the time, people just ducked their heads when I asked which I thought was kind of annoying. The people who did decide to respond were really nice about it and one guy even just walked up to me and stuck out his hand for a survey which I thought was really nice.

    I’m really glad that we shortened the survey because everyone who took it asked how long it would take and I said two minutes but it definitely took longer and I felt really sheepish making them fill my survey out. When friends filled it out, they would compare their answers sometimes after they were done which I thought was cute. Overall it was a nice experience though asking people was definitely kind of embarrassing.

  6. Vanessa Neergheen

    I was also part of the group that administered the survey online. For this reason, I do not have a particularly interesting story to share about the sampling process. Thanks to JJ we were able to use Google Docs not only to administer the survey, but also to collect the results. Can it get any better? As of now I believe we have 75 responses, though I am not sure if people from the other sampling groups have been inserting their data into this document as well.

    When I was sending out the interview I used my G-mail address, which is a little out of the ordinary and does not reveal my identity or affiliation with Middlebury (unlike our Middlebury e-mail addresses). At first I was concerned that people would be wary to fill out the survey, thinking it was a virus or spam. Once the individual opened the survey they would realize that it was associated with Middlebury, but they would have to at least follow the link. However, even if this was the case, did my strange e-mail address help make the process a little more random? I then started thinking about if/how our rate of response would change if we all used random e-mail addresses. Though it is important for our participants to know whom we are associated with (ethics!), can any sample ever be truly random?

    One question I had about the “Spin the Bottle” and mapping methods was how the presence of the individual administering the survey would affect the participant’s response. I can imagine that some people would be more hesitant to respond truthfully to a few of our questions if they thought that someone was watching over their shoulder. Did any of you “Spin the Bottle”/mapping methods people think this was actually an issue? Did you turn around or do something else to avoid watching participants as they completed the survey? Is there a certain protocol for this situation?

    At first I thought I would be using random sampling for my research project as well. I remember thinking as I was writing the “Methods” section of my research proposal, “Can’t go wrong with random!” However, after actually learning when it is appropriate to use random sampling, I think I will ultimately use non-probability sampling to choose my group of participants. I will be working with students from a very small private school, and I think it will be best to interview all the students. I may also work with students from a larger summer school program, and for this portion of my research I will probably use random sampling.

    On a side note, thanks to all of you awesome people who trundled about handing out surveys! You have good SOAN 302 karma coming your way, I promise.

  7. Elise Moody-Roberts

    I found the reading about sampling very interesting and confusing. I had to reread the various methods to fully grasp them and visualize how they work. At first, ‘random’ sampling methods didn’t seem random at all to me other than the effort behind it to make it random. Otherwise, it just seemed like a matter of chance or a crapshoot. But I guess that is what the intention of constructing something to be random turns out to be. In conclusion, I learned what ‘random’ really is.
    I was in the online survey group, and I have a great appreciation for Google docs now. Creating a random sample group and online survey can be technical and tedious work, but once it is done the efficiency of it is so rewarding – the responses come automatically already organized into the spreadsheet.
    I was shocked to see responses posted within minutes of sending out the survey. I imagined that students did it so immediately because it was something they could do and gain a feeling of accomplishment in a study break. I am also surprised at the response rate; it is much higher than I anticipated.
    The one thing I wish we could have done differently is to bcc the people we sent the survey out to. That way people would not know who else got the survey and would have less of an idea of who the research represents. Considering that everyone’s answers are anonymous, I suppose in the great scheme of things this does not matter, but it would increase the anonymity of the project.
    I am quite excited to analyze results, and also to hear about the experiences of the ‘map’ and ‘spin-the-bottle’ method groups – I am sure there are some great stories.

  8. Quan Pham

    Unfortunately my mornings and early afternoons are usually dedicated to other prescheduled engagements. I found myself to finally being able to start surveying this past weekend on Sunday during the early evening hours. Little did I know that the Vermont sun was going to set earlier on because of the recent Daylights Saving clock set-back. So at 5pm I found myself in the dark on Battel Beach. It’s obvious that I cannot expect my survey-ees to complete these forms in pure darkness, so I stood beneath a lamppost on the side of the walkway. As an unexpected student passed by I would quickly introduce myself to break his or her flow and ask for assistance with this project. I got about a half mixture of yes and the other half of no. I thought I was prepared — I brought a lot of pens with me that I know would have worked even if you were writing with it vertically (on the wall). But then again, when you are in the middle of the field there were no walls and I had no clipboards. In addition, 5pm on a Sunday was a peak study time so despite my high-traffic location, I had a hard time locating participants. I resorted to going inside Battel dorms and asking students who had their doors open. Right away I was able to accomplish this. I assume that if you are in someone’s comfort zone they would more likely to be helpful than in public. I then ventured off to Atwater and this was not much of a challenge because I found myself among friends, who helped by recruiting their friends.

    The next day, I needed to resume my surveying so I did so in front of the CFA. My actual location was behind the CFA, but I decided to tweak the plan just a little because I wanted to stand in a place with higher traffic. There were several problems. The fact that I am doing this on a Monday meant that all my colleagues had already spammed the campus. I had a hard time identifying an individual who have not yet participated in this survey. What’s more, 3p on a Monday afternoon is a busy time for athletes or students in transition. But when are students really free to stop and give me “3 minutes”? Yes, I did promise to my participants that it should not take them more than 3 minutes, and many would ask me to guaranteed that because they know that I was just saying what I needed to say to get away with it. Again, the lack of clipboards made some participants kneel to the cement sidewalk to fill out this survey. Also, at the intersection of the CFA, there were a lot of cyclists. I found myself to once be running alongside a cyclist just to get her attention and ask for her participation.

    On my next site near Prescott, I had to walk with a student as he headed to the gym. I walked with him for about the length of two football fields. This way he could continue on his journey and provide me the assistance I needed. Nonetheless, after receiving his survey, I just had to walk back to my original location. In addition, at this time during the day, the traffic around Prescott was minimal to none. I sat outside of the Prescott House for about 15 minutes before I saw someone and then after that I decided to move to the back of Proctor instead because it was getting a little ridiculous.

    My reflection on surveying includes the understanding that it is a very difficult process because you had to deal with people directly, make sure that you had all the materials you needed, occasions of participants already being surveyed (especially on a small liberal arts campus), and identifying an appropriate time when people are free, which may be never. I think that I cheated with the randomized experience in that I took to comfort asking those that I knew. When there was a complete stranger, I always had a five second debate whether or not I would disturb that person, but with a friend I instantaneously asked for his or her participation.

    Next comes the data input. I created an extensive excel sheet which quantified all our qualitative responses. Inputting each survey was quite tedious and I cannot imagine if one would have to do all 600 by him/herself. That would be a lot of redundant work; but I ensured optimization with the numbering system so that the data-inputter would not need to press anything more than a number and type, for the most part.

    Surveying was fun, but in the future I might consider doing it with a partner so that it is not as nerve-wrecking to break down that initial barrier when talking to a stranger.

  9. Danielle Berry

    I was part of the online survey group. I’m not entirely sure how many responses we got because our spreadsheet is shared with everyone else, but I think we got a little less than 50% for a response rate, which isn’t too shabby!

    I found the process of selecting a sample for an online survey to be very simple, especially since we had the directory to work with. Sebastian figured out how to copy and paste the directory into Excel while others used a random number generator to produce 200 random numbers between 1 and 2400(I forget how many students Middlebury ACTUALLY has, but it’s close to this). Sebastian and I discovered that incoming febs were included in the directory, but I knew how to sort in Excel so all the students without a commons (the new febs) would be grouped together, so we could delete them. Sorry, new febs! Maybe next year.

    Next, we tediously highlighted the cells/names corresponding with the random numbers Maureen had. I bet there is a better way to do it than we did, because it took a while. We deleted all the non-highlighted names and divided the highlighted names amongst ourselves and e-mailed about 30 students apiece. I think this was a good idea because there is no way to BCC the way we did it, and being on a list with 30 other students would make me more likely to respond than being on a list with 200. I would feel like my response was more important since I was only 1 of 30!

    I’m not exactly sure how JJ crafted the survey, but it looks awesome! Overall, I’m a little disappointed that I didn’t get to go stand somewhere random or knock on doors, but I am glad that I got to help figure out something technical. I think this will be a useful skill to have.

  10. Kyle Warner

    I still am not sure what exactly to make of sampling. I understand the ideas behind having a random sample, it many ways it ensures that you collect information from a variety of sources who differ in numerous ways. However, I am not sure how some of these methods were developed. How exactly a leading group of anthropologists, presumably with PHDs decided that spinning a bottle in the center of town was a scientific method I really don’t know. I feel like there are much easier methods of randomizing a group, and well much more sensible methods. However, it was still an interesting and worthwhile experience, despite the fact that I was constantly envious of the random email sample group.
    I was part of the random number generated rooms group, and it was relatively annoying. I feel like it might have been easier to just ask everyone in a building to take the survey instead of randomly selecting rooms. The main problem that I dealt with while handing out surveys was when people were not in the randomly selected rooms. This caused me some problems because I could not just go to their neighbor and have them take the survey, I would have to go through the process of reassigning a building’s rooms with numbers and using a random number generator to select me a new, and still random room to survey. While I would not say that this method was difficult it was in some ways just irritating. The whole random aspect of the surveys added a lot of extra work.
    However, I understand the ideas behind random selection and why it must be used. It is important to use random populations when research requires it. Random sets of populations allow for a better and more complete sample, a sample that could not be recreated without randomization. Randomization allows for a broad spectrum of respondents. This allows for a survey to not have an inherent bias that could possibly compromise the data. With data that is not truly suitable for the production of research, hypotheses and conclusions may be completely wrong for the question at hand. Moreover, it is also possible that the right question is never asked at all. Without the proper survey group the research is a waste of time.
    Despite my jealousy for the email group and the annoyance of unwilling survey subjects it was still an interesting experience that taught me a lot about interacting with a study group and also helped me see how other people saw a survey that I was part of creating. The ability to interact with a group that was not included on the actual research end of the project allowed me to see the problems that still existed within the survey that we created as a class, and it taught me a lot about how to interact with test subjects so as not to make it an annoying experience for them.

  11. HiMi Kanaumi

    I was a part of the Map Method group, and had randomly been assigned Hepburn front, Weybridge, Homestead, and the athletic center. Before collecting these surveys, I thought it would be really easy to get the 28 I needed so long as I was standing in an area that students walk by a lot, like the dining hall or the dorms. I also didn’t think there would really be a potential for bias. But what I noticed during this lab is that when you have almost 20 people walking back into Hepburn at once, it’s hard to get everyone’s attention. A lot of people were ‘in a rush’ and refused to take the survey, and so I usually called out people I recognized/friends. When I was standing alone, it was hard to get random strangers to take the survey, but when I was standing with five other people taking the survey, people seemed interested and were more than willing to take the survey. I guess my timing wasn’t the best either, since it was around 3pm that I started giving out surveys. People were late for class, had to go to practice, etc…

    Hepburn front and the athletic center were easy places to quickly collect the 8 surveys that we needed for each location, but Homestead and Weybridge, which are houses located right next to each other and are quite far from the rest of campus, proved to be a little more difficult. I actually kiiind of cheated (?) and took surveys as I was walking towards Weybridge/Homestead area. I didn’t exactly remember where the boundaries were so maybe it wasn’t cheating. I guess that’s one thing I learned throughout this process- know your boundaries!

    I feel like the method of going door to door is less biased. I think a good portion of the surveys were done by my friends/classmates/acquaintances, and this wasn’t even really intentional. Sometimes total strangers have no interest/incentive in participating in things like surveys, which is completely understandable. I promised myself that from now on, if someone asks me to fill out a survey and they’re standing in the cold, I’ll do it.

  12. Amy Donahue

    The beginning stages of random sampling via the mapping method were fun. As we drew line after line on the map of Middlebury’s campus, we found that using a random number generator to determine our cross-sections was actually quite frustrating at times. Our map became broken down into a bunch of tiny sections on the seemingly low populated edges of campus (recycling center, admissions, etc.), leaving huge sections centrally, that encompassed some of the areas with the highest traffic pattern (from Proctor to Old Chapel). While some of the group felt we should stick to the random number generator strictly, and adhere to the areas we were assigned to, others felt that since many of us were – at random – assigned to less frequented areas, it might skew the results and therefore would be better to regenerate a number. This raises the question- is the strict adherence to the methodical and structured process of randomization ALWAYS effective?
    The areas that I was technically assigned to were as follows: BiHall’s parking lot, outside the Admissions office, behind the athletic complex, and the back half of the recycling center. Due to nothing other than time constraints, I admittedly found myself having to fudge these geographical boundaries a bit. I crept into the mods parking lots, which provided a bit more student action than the recycling center on a Thursday afternoon. And during the time that I was near the admissions office, passersby tended to be more faculty and prospective students than the the members of the student body that I needed to be reaching, so I shifted slowly more and more toward the library for a bit more traffic.
    I found that the closer to a building someone was, the more likely they would be to stop and take our survey. I felt that when interrupting someone in the middle of their path to somewhere, they were more anxious about the time constraints traveling to their out-of-eyeshot destination and therefore more likely to refuse taking the survey. But feeling secure about having already arrived to where they needed to be, people were more available to give a few minutes of their time. The most likely candidates for a definite “yes” to filling one out definitely tended to be who were sitting down; almost all of these were more than willing to break from whatever they were doing – working or conversing – to take the survey.
    Concerning the content of the survey, many seemed concerned with the question that asked, “How much work did you do yesterday?” They would either point out to me verbally that “yesterday” was a Saturday (to excuse their notated workload being lighter), while others felt the need to indicate on their sheet that the great amount of time they spent working the previous day was due because it was Sunday. Perhaps instances like these should inform the way in which we proceed to fine-tune our survey.
    Overall, the experience was fun. I found it harder to approach people than I had expected it to be for myself; I chickened out at the last second a few times. But mostly, people were friendly and receptive and I found myself loving the moments of friendly exchange between myself and the survey takers. It was also interesting, in the end, to be plugging in the data I collected and notice certain patterns among them, and curious to know more about some of the people I had given a survey to.

  13. Isabella Tudisco-Sadacca

    I never knew random sampling could be so structured, or calculated. I always thought that when people were randomly selected it was less intentional than it actually is. I realize now that even in doing what we believe to be random and selecting people that cross our paths etc there actually can exist bias without us even realizing it. I was in the focus group that random sampled the dorms. We got very lucky in having the dorm information (room numbers and population per dorm) sent to us, without this it would have been much harder to calculate. The process of selecting the dorms was not necessarily that difficult but was time consuming and there was no real way to share that task as it involved mostly one person on a computer and maybe the rest of us adding numbers. What I didn’t realize would be hard is after we generated the dorms we would be going to, when you have a calculated outline of where to randomly go what happens when you cant get everything you need from those places, is it still random if you stray a bit?

    Carly and I worked together in handing out the surveys and the dorms we had to go to were freshman, mostly sophomore, and only senior dorms. We faced a few problems throughout the process that showed us we knew a bit less than we thought. We started in Coffrin, a dorm of mostly sophomores and were fairly lucky with the amount of people that were in their dorms, however we struggled with knowing what do when we handed out surveys and awkwardly would just wait in the hallways. When less than seven people were in the hall we went to we didn’t know what to do, if we should go down a hall and pick the next door, or just not complete all seven needed for the cluster. As Carly mentioned we had to regenerate a floor to get all that we needed in Coffrin. Most people were friendly, however a few people were a little cold yet still didn’t refuse to take the survey. Someone actually wanted us to hang out with them while they filled out the survey.

    Atwater was by far the most difficult, to begin with each building has three entrances which was not accounted for when we generated floor numbers but worked to our benefit as only four people live on each floor and being full of seniors, it was less likely that all of them were there. So we went to the same floor on each of the entrances. In this case especially I think what Bianca said is true about bias in terms of who is in their dorm at that time, we also found a lot of people who would just ignore us or were running about and so we couldn’t catch them. We encountered a few strange people in Atwater who were not the most pleasant to interact with but overall received the greatest diversity of responses from the Atwater suites, or perhaps just people with more developed ideas of work and play and more complicated schedules as they have been at school longer and are generally more involved people.

    Battell was flooded with people, which I suppose is the nature of freshman dorms. However a few people that filled out our survey were sick, hence why they were home, and therefore would always tell us that this week is not the best example of their work habits because they haven’t been able to do any work since they’ve been sick. I was a little shocked these people kept their doors wide open, a little inconsiderate perhaps? Also I think this shows a difference in age/year in that when upperclassman are sick you are less likely to find them lounging around in their room or not doing work as being sick becomes something you just deal with.

    I had fun with this exercise and think it was great way to be introduced to the survey process as we had a pretty good chance of high success. It was nice to be able to work with Carly because we were able to ask each other questions about the kinks that came up and it didn’t seem as tiresome to wait around on the steps of Atwater when there was someone else with you.

  14. JJ Janikis

    While I did not have the unique opportunity to gather a random sample of students with the “Spin the Bottle” or “Map” methods, I got to explore my technical side and create the on-line survey. With my love for Google Docs I was able to create a “Form” for the survey. Once the class decided on the final questions, it was easy to format the survey using the Google Docs survey options. Therefore, I could choose to create a multiple choice, checkbox, or fill in the blank questions. I separated the questions by type: background, work, and play on to separate pages. The survey was quick and easy to follow along for participants and only took approximately 5 minutes to complete.

    After testing the form within our group, we sent out the survey to randomly selected group of 200 Middlebury students. The random group was created using by putting the Middlebury student email list in an Excell spread sheet and randomly sorting the names. The first 200 hundred names were used for our survey. Names of students in the class were deleted to remove potential bias.

    The most exciting part of the Google Docs form is that once a student participant takes the survey their answers are automatically placed in an spreadsheet. The answers remain anonymous and are sorted by question. The form and spread sheet was shared with the rest of the members of the group and later shared to the entire class as a means of updating all forms of student sampling. Thank you Google Docs!

    I am glad that I was able to experiment with this new means of sharing information. Beyond this survey, I think this is a valuable skill in collecting and processing information with and for a a group of people. However, I still feel like I do not fully grasp the concepts of random sampling via the “Spin the Bottle” and “Map” clusters. How do these methods work when the sample is in an isolated location where perhaps there is not a lot of human traffic and interaction?

    While everyone was out waiting in random locations (i.e. Mods parking lot) and knocking on strangers doors, I was able to sit back comfortably and watch my participants fill out the survey live with their responses on the spreadsheet. (Sorry folks….) I look forward to seeing what the different biases will be like in the various sampling methods.

  15. Charlotte Gardiner

    We continue to explore the work/play dichotomy across the Middlebury campus, and this week we have employed sampling as our method for conducting research. I was placed in the Internet/online sampling group, so unfortunately do not have as much of an experience to report on as those in the other two sampling groups. We simply crafted a generic email and sent it to 200 randomly selected students (based on a random number generator) and attached the survey link to the email. We followed up with our selected individuals, and are still awaiting replies. Google Docs is a great program, as each member of our group can see the responses coming in and begin to analyze what the results actually mean.

    Currently, a little over half of the students have responded to the survey, which I feel is a respectable number thus far. I hope we can continue to garner responses, especially as we all just sent a second email reminder to the students. One of my friends was ironically on the group of students that I emailed and she sent me a personal email, and said her computer was broken, but that she would use a school computer when she found the time, so that she could respond to the survey. This made me question the “randomness” of this specific research method. It is funny to think that despite the fact that this is a deemed random sample, there are still ways for people to know who is administering the survey. Is this a problem?

    After reading the posts from other people, it seems that the “map” method worked effectively, and resulted in a high participation rate (perhaps thanks to the short length of the survey). People can opt to delete their email and as researchers we would never know if specific individuals from our random sample did in fact respond to the survey, but face-to-face confrontations make it difficult for people to escape and “delete”. I believe people would feel they were being rude, unless they were in a hurry, if they failed to respond to the survey when asked in person. On the other hand, the “housing” method seemed to result in “twisted” responses because ideally after freshman year students often choose to live together and therefore may engage in similar “work” and “play” activities, so I question how varied the responses are and how informative they may be if they people surveyed are so similar?

    Nonetheless, I do think random sampling is an interesting way to gather information because people ideally respond in an unbiased and honest fashion, especially through the Internet/online survey. Researchers can then truly evaluate the responses. However, while our survey is ethical, does it prove problematic when, as Internet survey-givers, we can see who we are sending the survey to (because we may know the individuals) and so can therefore technically interpret the information and discover whose response is whose?

    It is also interesting to note the advantages and disadvantages of random sampling as compared to non-probability sampling. Often times (which was technically not the case here), people don’t know you in random sampling, so this can affect the quality of responses in positive and/or negative ways.

  16. Maureen Wyse

    My group used an online survey to get responses. We first went into the directory and got a list of all the students then assigned numbers to them and selected 200 with a random number generator. This part of the process took the most time, since once it was completed we could email the survey out and wait for responses to come in. Within a day or two we were up to 50 and now I think we have 100 or more which is pretty good for a survey response rate.

    I think a big reason as to why we are getting a lot of responses is because the survey is so short. Also work and play may be a topic people are interested in since it affects everyone on campus. Everyone at Midd has experience in the topic and has a connection with it so more people may be willing to respond.

    Our group seemed to have it pretty easy since everything was online, but it would have been interesting to interact with the participants to see their reactions to certain questions, but using an online survey is nice because it keeps everything organized and it’s easy to use. The hardest thing our group had to figure out was how to eliminate incoming febs from the directory list within exel.

    We did send out a follow up email so hopefully our response rate will get higher over the next day or two.

  17. Amy-Leigh Scanes-Wolfe

    At first, I was very apprehensive about my ability to actually get fourteen interviews in the vicinity of the Mods and the recycling center parking lot. I was right to be apprehensive on some level. The first time I went out, it was bitterly cold, windy, and a total of six people strolled across the parking lot in an hour. However, it was all about timing. The next time, I got six people within half an hour. And in fact, my obscure location was a very helpful tool in getting cooperation. Most people felt so sorry for me that they agreed, especially the first day, when it was so cold.
    It made me think about the importance of timing. The parking lot is a great place on a Sunday afternoon, but less so on a Monday morning. Also, when I was surveying people near Munroe, I got great responses right after class let out, but less a few minutes later, when people were rushing to their next class.
    In this location, though, people tended to be migrating in pairs or groups of three. Without the enforced formality of a classroom, table, or chair, pressing their fluttering surveys against car windows, the aura of confidentiality and objectivity faded away. Within the group, there was a lot of joking and looking at one another’s answers. In one case, two of three guys who started the survey together finished long before the third. They proceeded to lean over and help him select the “correct” answers, all the while hassling him for treating the survey like a test.

    I also ended up getting one extra survey in the parking lot, because I approached two people together and felt strange making one wait while the other one filled out the survey. Also, it wouldn’t be random to select just one of the girls, since I saw them both at the same time.
    I also ran into the problem of duplicate surveys. On the cold day, I resorted to knocking on the doors of the Mods, which were just outside the border of my allotted section. It seems that other people assigned in that vicinity had done the same thing. Another guy was part of one of the other sampling sections; someone came to his room just the other day.
    Some people also took issue with the question asking, “How much work did you do yesterday?” On Friday, a bunch of people said, “Oh, but I had a midterm today, so…” and then on Sunday, a bunch said, “Well, yesterday was Saturday, so…” Thus while these specific questions are helpful for our knowledge, hoping the average turns out alright, the participants themselves feel uncomfortable answering them. It also goes to show that timing is important. If I did all the surveys on a Friday in mid-term season, it would be very different to a Sunday at the beginning of the year.
    All in all, though, the process was actually pretty fun. There’s something satisfying about doing the unexpected and seeing how people react.

  18. Carly Shumaker

    My group used the “Spin the Bottle” method in order to randomly sample students in the dorms. The process at first felt tedious when we added up the number of beds in each dormitory, found the total number each dorm could potentially sleep, divided by 200, used a random number generator to find our beginning point, and then repeatedly adding twelve until we found every dorm that we needed to go to according to populations of each. Just reading about it is tedious, no?

    After we got all of that sorted out, however, I had a lot of fun actually going to the dorms and pestering people to complete the surveys. Bella and I went to Coffrin South and Coffrin Center, Battell North and Battell Center, and Atwater Suites A and B. Like the others, we had some trouble finding people who were in their rooms between 2:30 and 4:30 p.m. Bianca made a good point when she said, “There might also be some bias to the people who were at home during these hours. Perhaps the people who are at home are generally lazier, whereas those who work and play more were busy running around.”

    In Coffrin, which is full of sophomores, we had to regenerate a floor randomly in order to find seven people to complete the survey. We definitely knocked on every single door in order to get seven students. We were generally met with positive attitudes. Many people were excited to take the survey, and if not excited at least very willing. One girl had already taken it online and one boy had taken it in the library, but they both were happy to take it again.

    Atwater Suites proved to be slightly more difficult, as one can imagine. The internal architecture of the building most likely lead to some bias because friends generally live together in this type of housing. In each building, we had to go to all three entrances just to find seven people. Many less seniors were at home at this hour than sophomores. One boy turned us down because he simply didn’t want to complete it. Another boy had a paper due in 15 minutes, so he couldn’t take the time. In the very last entrance of the last building, no one was home, however, we found two people standing in the building’s entrance. Although they didn’t live in Atwater, they gladly filled out the survey for us.

    In Battell, we had absolutely no trouble finding freshmen. It seems as though they all hang out in their rooms around 4 p.m. Many freshmen were there from other floors visiting their friends, but we had them fill out the survey, as well. They all seemed eager to fill out the survey, even though two girls were quite sick! Only one boy refused, saying he was napping.

    All in all, I thought this was a really fun exercise and I felt successful regarding the number of surveys completed. I was surprised to see the overall willingness to help out other students and fill out the survey.

  19. Jenna Reichenbach

    The “Spin the Bottle” method of sampling was a fun way to determine which dorms/rooms our group would be surveying; however, I don’t think that it was particularly affective on a college campus. Using the random number generators was easy enough–we were able to pinpoint the exact 30 dorms our group would be going to, then the floor of each, and finally a room number to start with. Unfortunately, many college students have a plethora of obligations! While traveling from dorm to dorm, it was incredibly difficult to find at least seven people to survey. The starting room was never occupied and nor were many others on each hall. I inevitably had to knock on most doors of each hallway before finding enough people to take the survey. In one building in particular, I went to all four floors in order to find enough people to survey (and even then I had to throw a survey into the face of someone walking into the building’s door on my way out).
    I think this method may not be the best to use on Middlebury’s campus, especially depending upon the time of day. Surveying in the late afternoon means that I didn’t find many athletes or students involved in afternoon organizations. I also surveyed many more females than males (although I’m not sure what this means…). It was an effective method in that individuals seem to have a difficult time saying no when you’re asking them in person. A few individuals hesitated and ultimately took the survey. Only two people turned me down as they were on their way to class.
    The most challenging part of surveying was when I moved on to dorms that weren’t of my class year. The first three dorms that I visited were all sophomore dorms, so I felt very comfortable knocking on the doors of acquaintances and people that I didn’t know, but who are the same age as myself. Once I moved on to Starr, I was much more nervous and intimidated by the fact that I didn’t recognize any of the names on the doors. After finding no more than one person on the initial floor selected, my anxiety increased and I felt very uncomfortable knocking on the doors of seniors. Almost everyone was very friendly, but I still felt that it wasn’t my place as a stranger to be interrupting their afternoons.
    At least four of the people that I surveyed had already completed the survey once or twice (online or in person just a few minutes beforehand), which means that we may not have gotten a great variety of people. In the end, I thought that asking people to take a survey in person was extremely effective, however, it may not have received the greatest response based on time of day or the dorms visited.

  20. Bianca Giaever

    My group used random number generators to find random clusters of rooms in dorms to survey. However, since housing isn’t assigned randomly, each of these clusters probably carried their own bias. Three out of my four clusters were entirely Freshman girls. This is because freshman halls are often divided by gender. Since we took the gender question out of our survey we will never know whether this had a huge impact on our survey, though I’m sure that it probably did. The cluster of boys I surveyed were a group of friends who chose to live together. They all play sports and joked about how they did very little work. There might also be some bias to the people who were at home during these hours. Perhaps the people who are at home are generally lazier, whereas those who work and play more were busy running around. In fact, the bros thought they were being targeted by the survey because they were so lazy. It turned out that randomly generating a room number on each floor didn’t really matter, since I pretty much had to knock on every door before I could find seven people who were home. Everyone was pretty enthusiastic about taking the survey and I didn’t have a single person refuse or say they’re too busy (although, perhaps all the busy people were out on the town). On the entire 5th floor of Stewart I was only able to find 6 people who were home (it’s a very small floor) so I had to regenerate a floor and room number and ended up with a boy on the fourth floor. One person I asked had already taken the survey another way (I think she took it online.) I did think this system worked well because we got people to take the survey who normally would never ever take it (ex: bros who don’t leave their suite, shy freshman girls). In Allen there were a ton of people hanging out in the Common Room and I was tempted to just get them to fill it out. However, I knew that wouldn’t be random. I thought about calling out the room numbers in my cluster to see if any of those residents were in the Common Room, but it became too difficult.

  21. Toby Israel

    The Map Method– Discussion Post
    This was a surprisingly painless process. At my first three locations I had seven people within minutes. At my last spot, in the woods near Brooker House, I didn’t see anyone for a while, and it was starting to get cold, so I started to wander in the direction of Proctor until I came across seven willing participants. That might technically be against the rules, but it was still completely random, so I figured it was okay.

    In the course of the afternoon I was surprised to come across several students who had already been surveyed, and an even more surprising number of alums. I also experienced a completely irrational sense of loss when I would walk past people on my way from one spot to another and think that they would not be represented in our data.

    At one point I think I may have acted in a manner that might create compensation bias. I was standing outside Atwater suites, and when I asked a student if he was in a hurry and he asked me if I had a key card, I offered to let him into the building in exchange for filling out my survey. Ethical enough.

    Our survey got quite a few chuckles for the explanations offered for why students might miss class or consume alcohol, which confirms my belief that including a bit of humor in a survey will probably work to the researcher’s advantage. I know I would rather take a survey that makes me laugh than one that bores me completely.

    I also found that almost everyone I approached was willing to fill out the survey. Except for a few that were rushing to class, every student stopped and took it—I would estimate an 85% success rate. This is definite evidence for the benefits of administering a survey in person, rather than sending out emails. Thinking about methods for my own project, I am realizing that I will probably get a lot more positive responses if I approach people in person rather than by email; it’s just much harder to turn someone down when they’re standing in front of you, especially if it’s cold out.

    I am not looking forward to entering the data for 28 surveys on excel, however.

Leave a Reply