Week 6 - Kaitlin

 Hi everyone, 

Bias is one of my favourite research topics to discuss and I really enjoyed reading your thoughts and insights about bias in your own research. I hope to continue this conversation in my post, by going over biases that could occur in my own research. 

A key part of my research methods is measuring type of contact each participant has with individuals in their social network, as well as their perception of social support with these individuals. Because I want to measure if there is a change in social support perception when contact is more digital vs in person, I proposed a data collection method, where participants would answer questions about contact type and social support before and during the pandemic. 

Because this study would be starting after the pandemic began, I'm concerned that participants may be subject to recall bias, especially when asked about their relationships before the pandemic. One way I'm thinking of decreasing this chance of bias is by formulating these questions to have categorical answers so participants do not need to remember specific details. However, I am a but stuck on this so I would love to hear any feedback or suggestions! 

Comments

  1. AdenOctober 25, 2020 at 12:15 PM

    Hi Kaitlin,

    I think that your question raises some interesting points. I think we honestly do not know how the pandemic is going to affect interpersonal relationships. I know personally that there are friends who I just have not kept in contact with over the past few months because it is so much more difficult to see them. It would be innteresting to be put to the test for how I remember these relationships.

    Maybe one way that you can specifically control the issue is to make sure that you personally do not provide any hints as to where you want the study to go. If you are doing the interview in person for instance, make sure that there are no obvious clues to your own opinions on the outcome of the study. I also think you are definitely onto something is using categorical answers. Creating comprehensive questions and answer options may take some of the burden off of the participants to remember specific details.

    I am not sure there is a perfect way to overcome recall bias because human memory is inherently flawed. Howerver, there are some steps to take to minimize the issue as much as possible.

    Aden

    ReplyDelete
  2. KaitlinOctober 29, 2020 at 7:57 AM

    Hi Aden, I think that is an excellent idea, thank you for sharing this suggestion! I agree, the composition of the interview will be essential in ensuring that recall bias, as well as information bias is reduced in my study.

    ReplyDelete

Post a Comment

Popular posts from this blog

Week 11 - Kaitlin

 Hi everyone,  In preparation for the upcoming research proposal, I've been closely analyzing my current methods and reflecting on possible improvements. One improvement was provided by a TA, who suggest that I conduct my research in the form of a longitudinal study, with a study period beginning during the pandemic and ending after the pandemic and social distances measures have ceased. I think this is an incredibly good suggestion, as this type of study would reduce recall bias, as participants would no longer need to remember what kind of social interactions were occurring pre-pandemic.  As always, any additional thoughts, comments or questions are welcome!  Kaitlin  
Read more

Week 6: confounding bias and research methodology (Erina)

I found that last week’s assignment helped organize and formalize my research question and research design framework. What I noticed while I was writing up my assignment last week, however, was that I need to start fleshing out my proposed research method.   I intend to use a survey to gather information on student’s perceptions of the online course they are taking. One difficulty I am currently facing is that I lack enough statistical knowledge to understand the different statistical tools used in the journal articles I am reading. At the moment, I am focusing on reading up on studies that used small sample sizes because my sample size will not be large. I hope to get ideas on appropriate statistical methods this way. Also, this week’s lecture brought my attention to an important issue with my research question: confounding bias. Because I will be examining student’s experiences with online class, confounding bias will be a problem if I do not appropriately select my test...
Read more

Week 7 - Erina

One of the readings this week showed that context plays a role in whether internet users are comfortable having their information or content they post online used for research purposes (Fiesler & Proferes, 2018). I found I agreed with a lot of the respondents in this study (e.g. being willing to have researchers use your information if you can remain anonymous, being aware that you will not have full authority with whatever you post online). However, it was alarming to see that the academic field is still in the earlier stages of developing a research ethics framework when collecting data from the web and that most Twitter users were not aware that researchers use Tweets for their work. Related to that note, I realize I will need to obtain consent from the professor AND students before I start observing the student's level of participation in an online class. I will primarily use a survey to gather student's perceptions of their learning experience. For the survey, I will...
Read more