<p>(kind of tickled to discover that the thread I originated popped up as a featured one when I signed onto CC tonight!)</p>
<p>
</p>
<p>It should not bother you at all, unless you plan to send professionally incompetent e-mails to random professors.</p>
<p>If you identify professors you really want to work with, plan your travel in advance, send an e-mail that refers intelligently to the professor’s work and your interests, and attach a short CV with your undergraduate background and GPA clearly spelled out, then you will have no problem getting useful responses.</p>
<p>A non-response in that case likely means the professor just doesn’t have funding or space for another student.</p>
<p>A mass mailing e-mail from a student will definitely get ignored (“I am deeply interested in your excellent research” without mentioning what it is).</p>
<p>In the case of the study, the e-mail was so inappropriate that I would immediate assess a close to 0% of ever being a mentor to any of the students. So whether I chose to meet any of them or not may depend on a slew of more or less subconscious factors. Including, perhaps, not even meeting Jamal because if he was naive enough to send such an e-mail, he may be naive enough to think if I talk to him he has a chance, and if he then gets a rejection he may think its racist. Quentin Oldsmobile III may be more resilient to a brush off after a little chat, plus his dad may send along a big fat check.</p>
<br>
<br>
<p>No problem</p>
<p>But sorghum, while you individually might not have responded to such an e-mail under any circumstances, evidently there are many, many professors that did. Only, significantly more responded to people with certain types of names. </p>
<p>Xiggi (and others) as a thought experiment, what would you have said if 99 % of white males got responses and only 3 % of minorities did? Would you still say that it was a meaningless study because we can’t know why the minority e-mails weren’t opened or whether or not they had been seen? </p>
<p>Or, let’s look at another scenario that avoids some of the specific issues you’re raising with this study. Suppose I send a black woman into a realtor’s office to ask about the availability of homes in a predominantly white area. The realtor tells her there are no homes available. The next day, I send a white woman in, and she is told about a home for sale in that neighborhood. With that sample size, I can’t draw any valid conclusions. It is possible that the realtor is racist, but it is also possible that the home in question was put up for sale on that very day, that the realtor had overlooked it in his file, or even that there are multiple realtors working in the same office and some are better at their job than others. On the other hand, if I repeated this experiment 3,000 times and found the same thing in a statistically significant number of cases, I can draw an inference that race made a difference even if I can’t point to any individual encounter and definitively attribute it to racism. Do you agree with my reasoning there, or not? I’m just trying to separate your specific objections to methodology from a broader rejection of the logic underlying such studies. </p>
<p>Awakeningvenus, I still wouldn’t worry about the survey, because I think the consequences of this bias are fairly limited, even if you do have an ethnic name (if you don’t, it really shouldn’t worry you). First of all, while the disparity is troubling, the fact remains that more than half of the minority questioners did get a response even in the category with the worst disparity, which still leaves you with plenty of potential mentors to choose from. More importantly, however, the power of a stereotype is at its strongest when race/sex/ethnicity/etc is the only thing you really know about a person. Plenty of people who make snap, perhaps unconscious judgments based on something like race are still perfectly capable of revising their opinions once they get to know someone. Plus, recall that there are actually many places that actively seek out female STEM students and workers. A conscious goal of looking for talent among diverse groups should outweigh any unconscious tendency to overlook a vague e-mail from someone with a certain type of name. </p>
<p>You can design good (or bad) experiments to test whether realtors respond differently to race. However, even there an impact may be due to stereotyping rather than racism. It would be a very odd realtor indeed to turn away anyone who demonstrably had money to spend.</p>
<p>
</p>
<p>Until very (too) recently, this was a PROVEN MATTER, with realtors fined routinely.</p>
<p>Perhaps they wanted to preserve their clientele who they believed wouldn’t want to live in a neighborhood where members of other races lived or who sold houses to “those people.” But realtors were at least partially responsible for the enforcement of residential segregation. Period. No question. @sorghum, it appears that you don’t know much about the history of race relations in the U.S.</p>
<p>Well, sorghum, that last post demonstrated to me that you are a white male. As a white female who has owned property, sometimes multiple properties, for 35 years, I can tell you there are STILL realtors who don’t respond to female calls, even when the caller (me) has significant resources! I guess they think there is a man somewhere in the background who makes the decisions. . .</p>
<p>Maybe, rather than assuming I am a white male, you can remember that realtors behave dramatically differently nowadays to groups they used to ignore in the past. Chinese, Russians, will get the red carpet previously reserved for Arabs or Japanese a few decades ago.</p>
<p>I am perfectly well aware that clients may specify racial criteria, and some realtors are craven enough to give in to them. I wouldn’t say that is necessarily the preference of the realtor.</p>
<p>If they didn’t take steps to ascertain if the email was opened, they have no way of knowing if the recipients ever saw the email.</p>
<p>For all we know, the non white male names triggered the recipients’ spam filters. Sometimes a spam email slips through my filters, and at present it’s all WASPy names. If I look in the junk email folder, there will be some email from close friends of dictators in Nigeria, etc. Their names are not Thurston Howell. </p>
<p>I don’t know how the people who write spam filters do their work. I would not be surprised to find that nowadays there’s a certain formula as to the pattern of email that person receives. </p>
<p>I would want a good study to make certain the professors received and read the emails. At present, it seems the professors accused of bias are more sensitive to phishing. They may also have more effective technology. </p>
<p>apprentice: The realtor example has been done and proven that bias does exist in those circumstance. That, and the studies in which identical resumes were clearly favored for non-minority applicants seem to me to provide much more compelling evidence of bias than whether or not a professor answers an unsolicited email. When a study only reports their “regression estimated gap” rather than the actual data, it makes me skeptical. When a study does not set a level of significance and uses multiple levels and also reports results as meaningful whether or not they are statistically significant, it makes me wonder. When a report does not try to figure out why a certain minority group (in this case, Hispanic women, had NO significant gap, it makes me question the degree to which the investigators were trying to really understand their results. I certainly understand that investigators have a strong interest in seeing results from their work and reporting this gap is interesting. Extending the finding that a gap exists in email responses to the conclusion that professors are less welcoming or willing to mentor minorities does not seem well supported and requires additional study. As some of the comments on the NPR site said, it may be different if the Asian students had American first names, suggesting the bias may be toward the cost and challenge of taking on a foreign student, rather than a racial or ethnic bias. </p>
<p>Mom2and: those objections sound reasonable, and I frankly don’t know enough about statistics to really evaluate how much of a difference the factors you cite make, so I’ll defer to those who are more knowledgeable in the field. </p>
<p>I do know enough to recognize that some of the objections people are citing are bogus. Periwinkle, your racist spam filter theory doesn’t hold up at all. Spam filters work by picking up on the ISP, the mail client from which a message was generated, and key words in the content and subject line of the message. They don’t weed out foreign-sounding names. Given that, arguing that we can’t make any conclusions from the study because it is possible that profs had other reasons for not responding or didn’t see the e-mail in the first place is a line of reasoning that would invalidate pretty much every study involving human subjects ever. You can’t possibly control for every possible factor of your subject’s lifestyle, so the best you can do is control for those factors that one could reasonably infer might actually be germane to your results. So, you’re going to account for things like demographics, but you’re not going to be able to account for things like whether or not it so happens that proportionally more members of group A than group B like Skippy peanut butter rather than Jif. </p>
<p>No one has presented a single valid reason that over a large number of e-mails, significantly more people in one group than the other would have had overactive spam filters, been out of the country, had a a policy of never responding to such e-mails, etc. Those are the kinds of chance-based things that a large sample group eliminates as serious objections. If I flip a coin ten times, I may indeed come out with heads 8 times and tails only twice, even though there is an equal chance of both occurring. If I do it 100 times, I may even come up with heads 80 times, although it is far less likely. Over 3,000 throws, however, we would expect the disparity to be at the very least greatly diminished. Of course, it might not come out exactly to 1500 and 1500, but that’s where margin of error comes in.</p>
<p>One interesting factoid in the study is that public college professors showed very little bias, especially compared to private college professors. Not sure if that was because public college professors did not respond to ANY of the groups in a big way, or if they responded positively to all groups equallty. With any statistics-based study, there is the possiblity that your result is due to the chance. For some of the groups, this was as high as 5% to 10%, with others smaller. If you do multiple tests on the same data, your changes of having something differ by chance alone go way up. </p>
<p>Agree that spam filters in general would not differentiate by name. </p>
<p>
</p>
<p>This is exactly the point. The results weren’t anywhere close to that, although no one who read the “reporting” on the study would be aware of it at all. The presumptive victims class received well over a 50% response rate (although as Xiggi has pointed out, we can’t tell if the response was to “get lost” or not). Most of the “gaps” were quite small indeed. If the “prejudice” shown by that study is all that women and minorities have to face, they can relax and worry about more pressing concerns.</p>
<p>
</p>
<p>First of all, allow me to say that I did misread your previous post regarding the variance in the anwers between males and minorities. I was actually addressing a different issue. Problem with answering posts on a cell phone! </p>
<p>Regarding an exercise that’d show such extreme results, I’d say that the results could not be dismissed, regardless of the flaws in the hypothesis. Only a fool would dismiss such results. However, I still would think that the results would be enhanced if the study could add a “WHY” it happened, and this by qualifiying the lack of answers and creating sub-groups (by repolling the recipients) to identify the exact nature of the negative replies.</p>
<p>Regardless of addressing a 99 vs 3 percent replies, this is hardly what has happened in THIS study. And why it would have been interesting to find out more about both the polite refusals and the unresponsives. I fully understand that some view, based on statistical evidence, that the N was large enough to dismiss coincidences, but it remains that decreeing that all unanswered emails were based on recognizing various races is a rather farfetched proposal. And this because of --if needed to repeat-- a poorly constructed email and a very short notice. In a way, the negative reaction might have been prompted by a series of factors that go well beyond the race of the sender, and this ranging from undelivered email to mere negative reaction about the content of the email. </p>
<p>Again, is there any meaningful difference between NOT answering an email and responding with “Given the short notice, I have to decline. Best of luck to you?” And is such difference enough to suggest that there is evidence of … discrimination? </p>
<p>In the end, just as in everything, we all have the right to interpret studies and statistics with our own set of criteria. I happen to think that the results indicated a possible issue that needs a … better study and a better approach, and especially one that does not jump to conclusions of racial discrimination unsupported by flimsy “evidence.” </p>
<p>As someone who was born in Brazil and grew up here in the USA, I never experienced any type of discrimination like what is being described. </p>
<p>My first name sounds very American and my last does not sound ethnic, but I can see how perhaps things would be different if my name was Juan Gonzales, Dashanique Williams or Muhammad Abdullah.</p>
<p>I am not trying to justify discrimination but let’s be honest, people will make assumptions based on your name, your ethnicity and some of these stereotypes are based on reality.</p>
<p>Someone had mentioned something about certain groups that should be avoided. This many not sound too politically correct but Indians and Chinese people are two groups in particular, that at least in my experience will not hesitate to ask a tutor 100, 200 questions, send them 20 emails or call them at 2 am if they have to. </p>
<p>It must be something cultural but to a person who is American or Western, it is very rude and so annoying!</p>
<p>from the original research paper:<br>
</p>
<p>Secondly, I am surprised that they got as many replies as they did overall. Half the emails were sent at 8am asking to meet that day and the email itself seemed sort of spammy:</p>
<p>
</p>
<p>The generic “Interested in your research” is really not how you get a response and the person is only just applying to programs (e.g. email me when you get in and are trying to decide if you should attend), and you are looking for opportunities for yourself vs. opportunities to help me in my research area (e.g don’t make it about you…not yet anyway).</p>
<p>I know that everyone got the same generic email, but I am surprised that they got as many responses as they did.</p>
<p>I didn’t realize that there was no analysis of the type of responses people got</p>
<p>If all the responses to the Caucasians was “thanks for your email, call me when you are actually accepted” and all the responses to the minorities was “glad to hear from you! Let’s meet today after lunch” it would say something different than what the study concluded. </p>
<p>I’m sorry @bschoolwiz, but you seem complacent in your prejudice against indians and Chinese students. They don’t seem to ask more questions than the average in my opinion. One thing I do know is that the two countries produce more stem students than your country of origin, Brazil. You have anglicized your name, or are ethnically European and are male, thus haven’t encountered this bias, but stop defending it. There is bias, but as several people pointed out it is possible to overcome. Most of these private universities have Chinese and Indian male and female professors on their faculty. Sorry, haven’t seen many Brazilian though. There are middle eastern professors, Eastern Europeans etc. Most speak better english (with an accent) than the average joe sitting at home watching the flintstones. I am not a professor, but my dad is. Most of his grad students have been foreigners with an odd American here and there. Americans were getting jobs after their bs in engineering, where as the foreign nationals were going for graduate degrees and then finding better jobs …here. My brother got his PhDs in mechanical engineering from MIT. Most of his fellow students were asians, middle eastern or European. Sorry…no Brazilian. So maybe, if a professor received an email from an American student interested in persuing a PhDs, he may open and read it just because of the sheer novelty of the situation.</p>
<p>^^
Such an overwhelming presence and contribution of internationals surely confirms the bias and discrimination in mentoring in our tertiary education! </p>
<p>A tad contradictory don’t you think? Or perhaps this is similar to the discrimination at the elite undergraduates that has kept the overrepresentation of Asians to only 300 to 500 percent. </p>
<p>
</p>
<p>Asking a tutor multiple questions isn’t necessarily a bad thing. If anything, I’d consider that making the most of one’s time with a tutor…especially if one’s not sure about a given subject/topic. Speaking as someone who was a tutor…frankly that comes with the job. </p>
<p>As for sending multiple emails in a short period or calling them at 2 am or during inopportune times…the only students who tended to do that IME tended to be overwhelmingly White students from upper-SES families with some degree of entitlement. </p>
<p>Know of a few who ended up burning bridges with their Profs in college/grad school because of the degree of obliviousness they demonstrated on this front…and then whine about how “it’s not fair” even though many folks like yours truly tried warning them off pulling that BS beforehand. </p>
<p>BTW: I am a Chinese-American and I knew better than to call Profs at 2 am or disrupt them in the middle of family holiday dinners for non-emergencies unlike some of those classmates I referred to above. </p>