Skip to main content
Search form
  • 00:12

    JANE PALMER: Hello.My name is Jane Palmer.I'm on faculty in the School of Public Affairsat American University in Washington DC.I teach classes on research methods and on social policy.My area of research is gender-based violence,and I have a particular emphasis on methodological issueswhen conducting research on gender-based violence.I'm a former social worker.I worked in Chicago and Saint Louis for about 10 yearswith children and families.

  • 00:36

    JANE PALMER [continued]: So a lot of the work I do is informedby what I learned from those years as a social worker.In this tutorial, I'll be addressing two major questions.First, I'll be addressing the questionof what do we know about sexual assaultand other types of victimization on college campuses.And second, in light of recent conversationsinstructing colleges and universitiesto conduct surveys of their students,I'll be talking about what your college or university wouldneed to consider if they're planning on conducting a campusclimate survey on these issues.

  • 01:12

    JANE PALMER [continued]: There's been a lot of recent newsand national attention on the issue of sexual assaulton college campuses.The general estimate is that between 20 and 25 college womenwill experience sexual assault on a college campus.And this isn't an issue that's only limited to women.This is an issue that affects menas well on college campuses.That estimate of 20% to 25% has been consistently foundin multiple studies at a variety of universitiesthroughout the United States.

  • 01:38

    JANE PALMER [continued]: The only national data that we have alsofound the same estimate of 20% to 25%,but that national study was conductedin academic year 1996/1997, so that datais getting a little old.There are no concrete plans at this timeto conduct another national study.Instead, colleges and universitiesare being instructed, and in some casesmandated, to conduct campus surveys on their own campuses.

  • 02:01

    JANE PALMER [continued]: Maryland recently passed a law thatwill require all public and private institutionsin Maryland to conduct these types of surveys.Currently, if a campus doesn't conduct a survey,the only data they have to rely upon to really understandthe extent of sexual assault on their campusis from the Clery Act.The Clery Act requires all colleges and universitiesto report instances of sexual assaultthat occur in a given year on their campus.

  • 02:28

    JANE PALMER [continued]: And this provides some information,but it's limited information about the extentof the problem.Due to the way the law was written,the Clery data only includes certain sexual assaultsthat were reported to certain peopleon campus and sexual assaults thatoccurred in what's called Clery geography, which does notnecessarily include off campus locations.

  • 02:50

    JANE PALMER [continued]: And it's estimated that on some universities, up to 2/3of assaults happen off campus.So that Clery information is limited.Sexual assault is often not reportedto campus administrators or law enforcement or public safety.It's the most under-reported crime.Well, actually there's one crime that's more under-reported,and that's pick pocketing.

  • 03:14

    JANE PALMER [continued]: But other than that, it's the most under-reportedviolent crime.And so colleges don't have the informationthey need to really inform the types of prevention programsthey're offering or even their response to these crimeson their campuses.One way that colleges can get this informationis by conducting these online surveys.These online surveys can get information about victimizationthat a student has experienced or eveninformation from the victim's friends about situationsthat they've witnessed or the extent to whichthey see sexual assault as a problem on their campus.

  • 03:46

    JANE PALMER [continued]: Much of the conversation has focused on sexual assault,but unfortunately, dating violence and stalking alsoare problems on college campuses.I've looked at a review of studies.I've reviewed several studies, and there's an estimatethat 16% of college students experience physical assaultin a dating relationship.Up to 30% experience psychological abusein a dating relationship.

  • 04:11

    JANE PALMER [continued]: And there's an estimate that 9% experience sexual assaultwithin a dating relationship.The review of studies around stalkingis more limited than sexual assault and dating violence.But the studies that I reviewed found an estimate of up to 27%of students will experience stalking.And that could be in a dating violence situationor not that stalking could occur.

  • 04:35

    JANE PALMER [continued]: There's also recent attention being paid to cyber stalkingor cyberharassment.There's less research in that area,but those are the types of questionsthat could be asked in the surveyto understand the extent of the problem at your campus.

  • 04:55

    JANE PALMER [continued]: These types of surveys will really help youor campus administrators and faculty and studentsknow the extent of the problem on their campus.As I mentioned, the national data that we haveis old and needs to be updated, and institution-specific datathrough the Clery Act is very limited.And so this will help you know what's going on on your campus.

  • 05:16

    JANE PALMER [continued]: There are several considerations, though,before you go off to conduct this survey.You have to be thoughtful about whatsurvey questions you include on the instrument, how youadminister the survey, the impact the study might haveon participants, and the overall impact the studywill have on the college itself.So when you're thinking about the survey questions,you have a few options.

  • 05:39

    JANE PALMER [continued]: You can use an existing survey with, of course,permission from the author.You can adapt an existing survey-- again,with permission from the author of the survey--or you can create your own survey.And each of these options have their pros and cons.If you use an existing survey, what'snice about that is that you could have a comparison.So say I know somebody at another university that'splanning to do a survey.

  • 06:01

    JANE PALMER [continued]: If we use the same survey instrument,we could compare and contrast our different institutions.That could be an advantage.A disadvantage of using an existing surveycould be that it doesn't really apply to your contacts.So say I have a colleague at a public universityor a community college or a private university.There contexts-- the extent to which students live offcampus, the extent to which studentsengage in certain behaviors-- might be different.

  • 06:27

    JANE PALMER [continued]: And I might need a survey that is adaptedto my environment, which is the next option-- that youcan adapt an existing survey.Of the options, I would recommend this option,because you can use survey questions thathave been tested in other environments,but you can adapt them to your specific context.

  • 06:47

    JANE PALMER [continued]: And again, you'd need permission from the author,and I would recommend that you pilot test any changesyou make at your university to make surethat it makes sense to students that attend your university.The final option is that you can create your own survey.And this could be an option if you have a faculty member thatis an expert in survey methodologyand how you write questions right and well,and also has some expertise in the areaof gender-based violence.

  • 07:16

    JANE PALMER [continued]: There's been three decades of researchon how to do these types of surveys,so it's important to rely on that literatureif you're considering writing your own surveyor using an existing survey.This research that I mentioned hasfound that there are a couple of best practiceswhen you're considering how to word questions,especially around victimization.These surveys that you conduct, the campus climate surveys,can include questions about victimization.

  • 07:38

    JANE PALMER [continued]: They can include questions about bystander intervention.They can include questions on the university's responseto sexual assault.Ever since Professor Mary Koss's studyin the '80s using the sexual experience survey,there's been decades of research on how to word these questions.The first thing that we've learnedis that the questions should be behaviorally specific.

  • 07:58

    JANE PALMER [continued]: So they ask about behaviors that mayhave happened of experiencing unwanted sexual activities.And the questions don't specificallysay rape or sexual assault. They don't say were you rapedor were you sexually assaulted.Some studies have found that-- one study in particularfound that 50% of women who experiencewhat would be legally defined as rapedidn't identify it as rape.

  • 08:21

    JANE PALMER [continued]: There's some stigma attached to that word,and somebody may not want to identify as a rapevictim or a rape survivor.And so the way that the questions are wordedis really important to be sensitive, but alsoto capture the estimate of experienceswith unwanted sexual activities or sexual victimization.

  • 08:45

    JANE PALMER [continued]: So the survey instrument that I use was pretty comprehensiveand had questions on sexual victimization,sexual harassment, cyberharassment,dating violence, stalking, bystander intervention,and the university's response to sexual assault.And those are the primary areas that you'llwant to assess if you're doing a campus climatesurvey on your campus.

  • 09:06

    JANE PALMER [continued]: There are some great models online that you could use.There's one that was published by the Officeon Violence Against Women.It's a tool kit that was included in the NotAlone report that the White House releasedin spring of 2014.And there's also a pilot study going onat Rutgers University out of the School of Social WorkCenter on Violence Against Women and Childrenwhere they have a survey instrument that they'repilot testing that could be useful to youas you're developing a survey for your campus.

  • 09:35

    JANE PALMER [continued]: But as I mentioned, it's important to pilot testany survey instrument, any survey questionsthat you use on your campus.And when you're considering which survey questions to use,or if you're adapting existing survey questions,it's important to consider what you want your referenceperiod to be.And what I mean by reference periodis when you're asking about victimization,it's important to put a time frame in the survey questionof what you're asking about.

  • 10:01

    JANE PALMER [continued]: So you have several options.Some sexual assault surveys have asked about experiencessince the age of 14, or maybe said before 14 or since 14and asked for both childhood experiences and thenteen forward experiences.Some surveys have said since you turned 18.The campus surveys tend to say since youentered college or since the beginningof this academic year.

  • 10:24

    JANE PALMER [continued]: So when you're asking that victimization question--in my survey, for example, which is conductedat the end of March of an academic year,we'll say since the beginning of fall 2014,have you experienced this, that, or the other,however the survey question is worded.And that helps in a lot of ways.There's a couple things.Some research has found that a six-to-seven-month period isa good time frame for a survey respondent's memory to be ableto capture whether it happened during that time.

  • 10:54

    JANE PALMER [continued]: And if you're specific to instancesthat happened on a college campus--if that's what your concern is-- you want to be able to limityour questions to an academic year estimate,but there also might be reasons to ask aboutbefore this academic year or before they entered college.Because while sexual assault happens often or can happenon college campuses, some individuals come to campusalready having experienced sexual assault,and that's important information for a college to know,because those individuals will also need services--may also need services-- and may alsobe at risk for re-victimization.

  • 11:31

    JANE PALMER [continued]: So those are just some important thingsto consider as you're deciding which victimization questionsto ask, is what your reference period would be.It's important that the survey is self-administered.And what I mean by this is that the individual who'sresponding to the survey is taking the survey themselves,either on paper or online.

  • 11:56

    JANE PALMER [continued]: This adds to some of the confidentialityand also will increase that person's ability--it helps to avoid what's called social desirability bias.So if I'm interviewing you about a really sensitive topic,you might be less likely to sharewith me, a complete stranger, some really hard experiencesthat you've had.And so it helps if the survey is self-administeredwhere the student is filling the survey out themselves.

  • 12:20

    JANE PALMER [continued]: One, you could do a population survey.That's when you send the survey to everybody on your campus.And depending on the size of your campus,that might make sense.If it's a small campus or even if it's a large campus,it might make sense to send a survey of this typeto all students.There's a couple advantages of this,is that all students receive it.And so if you have a large campus,you could still have a pretty large sampledepending on who responds.

  • 12:43

    JANE PALMER [continued]: It also could have symbolic implicationsin that since all students are receiving it,all students know that you as a universityare concerned about this topic.But there are some disadvantages that I'll talk aboutin a moment when I talk about another type of samplecalled a random sample.If you don't do a population survey,another non-random sample would beusing a subset of students-- so all incoming first yearstudents, you could do a survey during new student orientation,or all students involved in Greek life or athletics.

  • 13:16

    JANE PALMER [continued]: And you might have specific reasonsto target those students to be askingthese questions, because of risk factors or that sort of thing.And that is really up to you of what makesmost sense for your campus.Another sampling method is random sampling,and this can really help reduce somethingcalled response bias where any survey will have response bias.You know when you get a survey in the mail or an email,some surveys you may respond to and some surveys you may not.

  • 13:40

    JANE PALMER [continued]: And there's always going to be response bias of whoresponds to a survey.And when it comes to a survey on a sensitive topic,this is especially true.The people that respond will be peoplethat are concerned about this issueor appreciate that you're doing the survey.If you do a random sample, this might reducethe extent of response bias.

  • 14:01

    JANE PALMER [continued]: You'll still have some response bias,but at least you're sending it to a random sample of people.You can also do what's called a stratified random sample,and that's what I do.I send the survey out to an equal numberof first year, sophomore, junior, and senior studentsto try and get a good cross section of the university.Some other campuses might stratify by genderand send the invitation to an equal number of malesand equal number of females.

  • 14:26

    JANE PALMER [continued]: And each of those have their pros and cons as well.And another consideration is whether you'regoing to provide incentives.And so this consideration goes in tandemwith the consideration about sampling,because if you're going to provide incentives--whether you're going to provide incentivesor what type of incentive you might providemay depend on how large your sample is.

  • 14:49

    JANE PALMER [continued]: So you could provide no incentive whatsoeverand just hope that people will, outof the goodness of their heart, fill out the survey.And that is fine, and people do that all the time.But it would help your response rateand it would help reduce response bias if you doprovide some sort of incentive.And there's a couple different ways to do this.

  • 15:11

    JANE PALMER [continued]: You may do a raffle.And this is pretty common, especiallyif there are budget constraints, wherea certain number of students who complete the surveywould be entered into a raffle to receivean iPad or an amount of money or an Amazon giftcard or something like that.And that does work.Studies have found that students do respondto that kind of incentive.

  • 15:32

    JANE PALMER [continued]: You could also provide an incentive--some studies have figured out ways with online surveysto provide a gift card to a businessafter they complete the survey.So if they complete the survey, they press Submit,they get sent to another website wherethey can enter their information and receivea gift card or some kind of incentive that way.

  • 15:53

    JANE PALMER [continued]: And that's another option.But with the raffle and the post-incentive--I don't use either of those options.I use something called the pre-incentive,and I'll tell you why.So I use a method called the Dillman Tailored DesignMethod by a survey methodologist named Don Dillman.

  • 16:13

    JANE PALMER [continued]: And as part of this method, he recommendsproviding a pre-incentive to all students,or everyone in your sample, whether or notthey complete the survey.And we do this for a couple reasons.One is that it really does help the response rate,because somebody already has the incentive in hand.It's not like, oh, maybe I'll win the raffle.There are some people that just never win raffles.So the raffle may not be an incentive for themto fill the survey.

  • 16:35

    JANE PALMER [continued]: But if they already have the incentive whenthey get invited to take the survey,they might be more likely to take it.It also helps us to keep who participated confidential.So this is a really sensitive survey.And if you do a raffle or you do a post incentive,you would have to know that somebody took the surveyto be able to give them the incentive.And so what I tell students is one of the reasons we--and is true-- one of the reasons weprovide the pre-incentive is that wecan keep who participated confidential,because that confidentiality is really important in this typeof sensitive survey.

  • 17:09

    JANE PALMER [continued]: So the different types of sampling that you use,the different types of incentives that you use,will affect your response rate.So population surveys-- surveys without incentives--may have a 10% to 20% response rate.The Dillman method, I've used it four times now,and other colleagues have used it as well.It's consistently resulted in a 60% to 60%response rate, which is really a strong response rate.

  • 17:35

    JANE PALMER [continued]: There's still some response bias.As I mentioned, there's always going to be response bias.But using the random sample and using this $2 incentive reallydoes help reduce the amount of response bias.So as I mentioned, you have an optionof whether you want to do an online survey or a papersurvey.And it depends on your campus what makes more senseor what is more feasible on your campus.

  • 17:57

    JANE PALMER [continued]: If possible, I recommend an online, web-based surveyusing a survey software that has securitybuilt in that you can confidentiallytransmit sensitive data.If you do a paper survey, you havesome additional logistical challenges.

  • 18:19

    JANE PALMER [continued]: It's more labor intensive than an online survey.You have to actually enter in the responses for all the papersurveys.And depending on your sample size,that might be quite labor and resource-intensive.The only case that I would recommend a paper surveyover an online survey is that if you have a captive audiencesituation.So say you're doing a new student orientation,and you've got 500 students in an auditorium,and you don't have enough devices-- iPhones, iPads,computers-- for them to do a web-based survey.

  • 18:50

    JANE PALMER [continued]: It's just not logistically possible.But you don't want to lose this captive audience.That would be a circumstance whereI would say a paper survey would be practical.But generally, a web-based surveywill cause less headaches, hopefully,and be a preferred method.

  • 19:14

    JANE PALMER [continued]: So I have a couple other issues that I just wantto touch on in this tutorial.One is there are some concerns-- I'vetalked to some administrators thathave heard about these potential mandatesto conduct these surveys.I've read articles as well in the newspaperabout this-- that there are concernsthat these types of surveys are too sensitive, that they'llnegatively impact students, that they may re-traumatize victims.

  • 19:38

    JANE PALMER [continued]: And so I decided to look into the literature on this,and I found several empirical studiesthat have found that, in fact, participantsin victimization surveys don't experience negative impacts.In fact, some really find it to be a positive experience.Because as I mentioned-- well, these types of crimesaren't reported.

  • 19:60

    JANE PALMER [continued]: We don't really talk about them much in society.There's a lot of stigma around them.And so to have this confidential opportunity for an experiencethat you had to maybe make an impact on your campus,on programs or policy, is really-- victimsreally appreciate that opportunity.There was one study that actually gave some studentsa sensitive survey on experiences with trauma,questions about sex, and they gave another set of studentsa survey testing their cognitive abilities.

  • 20:32

    JANE PALMER [continued]: And what they found was the students thatfilled out the trauma and sex surveyshad a much more positive experienceand felt that they were doing somethingpositive for the world and less mentalduress than those students that did the cognitive survey.

  • 20:53

    JANE PALMER [continued]: In light of asking these sensitive questions,there's a couple important thingsthat will help reduce the impact on participants.So as I mentioned, the sensitively-worded questionsfollowing those decades of researchon how to ask these questions rightwill definitely reduce any impact.Keeping participation confidential or anonymouswill also help not only elicit better responses in that itreduces social desirability bias if the person won'tbe identified, but it also will protect the studentsin the long run.

  • 21:24

    JANE PALMER [continued]: And therefore, it's important to get institutional review boardapproval to make sure that human subjects are protected.And it's also important to provide communityand national resources to participants in casethey are dealing with some hard stuffand want to talk to somebody about itor the survey brings stuff up for themthat they weren't expecting.It's important to provide in that survey in the materialsyou provided the participant some resources,some hotline numbers or online chat optionsfor them to be able to talk about anythingthat they might be going through.

  • 21:55

    JANE PALMER [continued]: And that helps reduce any impact that the studymight have on participants.And then finally, I wanted to addressthe impact of this type of study on the college or universityenvironment.As I've mentioned, I've done four or five of these surveys,and I've found that it really helpsadministrators understand the scopeof the issue on their campus.

  • 22:21

    JANE PALMER [continued]: As I mentioned, national data only tells you so much.It doesn't tell you what the student walking down the hallhas experienced.And it helps to inform policies and programsspecific to those campuses.And as one administrator said, no amountof stories from front line staff can do what numbers can do.And she's found that these numbers have reallymade an impact on the campus.

  • 22:43

    JANE PALMER [continued]: I've also found that students greatlyappreciate that administrators, that the campus,conducts this type of survey.It shows that the administrators are concerned and wantto be proactive about addressing this issue,and it also helps students inform some of their advocacyand activism as students in helpingmaking their campus a safer place for all students.

  • 23:07

    JANE PALMER [continued]: So I hope that this tutorial has been helpful to youin understanding what the decision points areif you're planning on conducting a campus climate survey.And I hope that you will look into someof those additional resources to help make your studythe best that it can be.


Dr. Jane Palmer explains how and why to conduct a campus climate survey about sexual assault. She describes what we already know about gender-based violence on campus, the benefits and drawbacks of campus-specific research, and the best practices for creating/adapting a survey to measure this data.

Looks like you do not have access to this content.

Conducting a Campus Climate Survey on Gender-Based Violence

Dr. Jane Palmer explains how and why to conduct a campus climate survey about sexual assault. She describes what we already know about gender-based violence on campus, the benefits and drawbacks of campus-specific research, and the best practices for creating/adapting a survey to measure this data.

Copy and paste the following HTML into your website