Connecticut College administrators responsible for investigating multiple reports of voyeurism occurring in residence hall bathrooms between October 2018 and January 2019 overlooked key digital evidence that College-contracted external investigators and the New London Police Department eventually used to identify the alleged culprit, a College Voice investigation has found.
Between October 17 and January 23, four female students reported five different incidents in which they believed they had been recorded in shower stalls without their consent. The second of the five reported incidents occured in the all-freshman Morrison House, while all other reported incidents happened in Plant House. Carlos Antonio Alberti ‘21, who police first interviewed on February 7 in his room on the third floor of Plant, became the primary suspect in the case after NLPD Detective Keith Crandall gained access to and analyzed residence hall card access logsand found that Alberti was one of two students who had used a student ID card to enter Plant and Morrison in the 12-hour time frames before each of the five reported incidents. Police documents indicate that 213 videos were found on Alberti’s multiple electronic devices that included recordings of more than 30 women at the College.
According to campus safety director Mary Savage, the College first shared residence hall access logs with NLPD on January 2, 2019 — 78 days after the first reported incident on October 17. New London police chief Peter Reichard would not say when residence hall access logs were first requested by or made available to NLPD. But, police documents that describe the timeline of the NLPD investigation indicate that the analysis that led NLDP to identify Alberti as the primary suspect occured in late January after the fifth reported incident. Title IX Coordinator Ebony Manning also told The College Voice that in late January, after the fifth reported incident, the College hired an external investigator and shared with them “more extensive” residence hall access logs. According to Manning, that investigator was able to find Alberti’s name in the datasets provided by the College.
Based on police documents, student tips, and interviews with College administrators and New London’s police chief, the College Voice investigation reveals that an internal Title IX preliminary inquiry, led by Assistant Dean of Residential Education and Living Sara Rothenberger and supervised by Manning, lasted for over three months before an external investigator was brought in. During the preliminary inquiry, Rothenberger and Manning used some version of residence hall access logs, along with Wi-Fi access point logs, to figure out who may have been at the scene of the different reported incidents, but never interviewed or involved Alberti, who lived down the hall from at least two of the four reported incidents in Plant.
Though police and external investigators found Alberti’s name in logs that covered incidents going back to October, both Rothenberger and Manning said that Alberti’s name was not on their radar. Both have said that they are unfamiliar with how the data they used was compiled, but also indicated that they were reliant on that data to guide their inquiry. The case raises questions about the College’s ability to conduct data-driven investigations of serial policy violations while utilizing accurate information or appropriate tactics.
“I absolutely didn’t feel like I was missing anything because I was going through everything”
In an open forum on December 11, 2018, Dean of Students Victor Arcelus noted that the College has the ability to use residence hall access logs during an investigation: “We have those records, so we’re able to access that information. It’s not something that we regularly track, we’re not watching where each of you are going. But it’s something that we can access if we need to for an investigation.”
Arcelus also described the investigation as “a community effort … we really depend on the community to help us gather the information that’s necessary for us to be able to get the details.” He said the administration would not be able to quickly resolve the matter “because there’s such limited information about who’s doing this, that there’s not a lot to go on.” Arcelus asked students for “vigilance, alertness, to let us know if there’s anything in any spaces.”
After the first reported voyeurism incident on October 17, Manning launched a preliminary inquiry and assigned the case to Rothenberger on October 23. According to Manning, the College uses preliminary inquiries in Title IX investigations to “figure out if there is enough to move towards a formal investigation, if there has been a policy violation.” She also said that it is standard practice for the College to bring in an outside investigator. Both editions of the spring 2019 Student Handbook indicate that preliminary inquiries “typically take 1-7 business days to complete,” but does not mention external investigators. The College’s most recent Title IX policy, dated to May 1, 2019, does reference external investigators but makes no mention of preliminary inquiries. The Student Handbook outlines parts of the Title IX investigation process, including preliminary inquiries, but makes no reference to external investigators.
But, according to Rothenberger and Manning, this preliminary inquiry lasted for three months before the College brought in an external investigator after a fifth reported incident of voyeurism occured in Plant on January 23.
Rothenberger suggested that the preliminary inquiry was extended after each additional incident: “This preliminary [inquiry] was several different things together … it was the same action but different people who received the action … so each time there was an incident, then there was another part of that preliminary [inquiry] that was started.” Prior to Alberti becoming the primary suspect in the case, Manning noted that at no point was it clear if there was more than one perpetrator: “Is it one person? I don’t know. Is it a couple of people? I don’t know.”
It is unclear why the case was Rothenberger’s responsibility for three months, or why Manning did not bring in an external investigator earlier. Though the student handbook outlines some details on how preliminary inquiries and formal investigations should be conducted, no guidelines address the possibility of an unknown perpetrator.
In the past, the College has not always strictly adhered to the investigative formalities outlined in the Student Handbook regarding preliminary inquiries and formal investigations. A February 25, 2019, letter addressed to college president Katherine Bergeron from the U.S. Department of Education Office for Civil Rights (OCR), obtained by The College Voice via a Freedom of Information Act request, suggests that the College has previously combined stages of the Title IX investigation process. The letter was sent to Bergeron informing the College of the outcome of a 17-months long OCR investigation into Connecticut College initiated by a complaint of sex discrimination during a Title IX case that involved alleged sexual misconduct between intimate partners.
“Based on OCR’s investigation to date, however, it appears the College may have processed the complaint filed by the Complainant as a combination of the first two stages [preliminary inquiry and formal investigation], which may have been intended to streamline the process in light of the evidence already gathered in the investigation of Student B’s [a separate student’s] complaint, but may have resulted in Student B not receiving notice about or the investigation of Complainant’s allegations, nor an opportunity to provide witness.”
Rothenberger identified herself as a “preliminary investigator” and maintained that she was not involved in the formal investigation stage; it does not appear to be the case that the College combined the preliminary and formal stages of the voyeurism investigation.
However, Rothenberger said that “the purpose of that preliminary inquiry is to see if there is enough evidence to go forward a formal investigation.” The College’s Title IX investigation procedures outlined in the handbook depend on evidence meriting a formal investigation. Though the student handbook outlines some details on how preliminary inquiries and formal investigations should be conducted, no guidelines address the possibility of an unknown perpetrator.
Rothenberger said her primary task was to “gather information about what had happened.” She said she interviewed willing complainants, referring to the women who reported incidents of voyeurism, and potential witnesses. Rothenberger said that “witnesses were gathered through Wi-Fi access data points that [were] from I.T. that were given to me and then door access data points that were given to me as well.” Rothenberger noted that a Title IX investigation “doesn’t go to a formal investigation until after a preliminary investigation.”
During the months-long preliminary inquiry, Rothenberger’s efforts focused on information relating to student cell phones and an attempt to figure out who was in the buildings at the times of the incidents. She said that victims had given descriptions of seeing an iPhone and a hand, which Rothenberger claimed were “the only bits of data we really had to go on.”
“In the interviews, I would ask students, regardless of gender, whatever, can I see your phone? They would put it out in front of me, and all I was interested [in] is the make and the color, because that was all the data we had.”
Rothenberger says she did not recall requesting to see the contents of student devices, but she did note that in one case, a student, whose phone shared some characteristics of the device described by victims, volunteered to show Rothenberger some of the contents of their device “because they wanted to show what they had been doing that night in question.”
According to Rothenberger, “anybody who was proximal at the time of the event happening based on [Wi-Fi and residence hall access logs] are people that were interviewed.” She said she did not speak with all residents of the dorms, claiming that based on Wi-Fi and residence hall access logs, “you could tell at the time who was in, or maybe not who was out, but you could tell who had accessed and come into the building at a particular time, and due to the timing of events, there was a certain subset of people that you could interview.”
Rothenberger said that, based on information she had in her possession, there were individuals who she did not interview. According to her, those individuals were “excluded based on time… I guess it was just a time exclusion that we did.” Rothenberger estimated that she conducted “roughly about a dozen [interviews] for each different incident… that went on between October and January.” She also noted there were some additional students who she interviewed who did not appear on any “lists” but whose names came up in other interviews.
Rothenberger said she never interviewed Alberti, but was unclear about whether or not Alberti’s name could have been on any list in her possession. She said that his name “didn’t come up, I don’t know… the case went to external investigators… before his name ever came up in the investigation.”
“At the time I absolutely didn’t feel like I was missing anything because I was going through everything,” said Rothenberger. However, she said that “there were many names on lists and so I don’t recall who was and who wasn’t [on the lists].”
“The data came to me…I didn’t talk with the person who ran the report”
Rothenberger said that she and Manning were responsible for determining what types of data should be used during the inquiry. She said they worked together “to figure out what information we should get… Dean Manning, as the Title IX coordinator, then asks other people for that information.” But, both indicated that they were unfamiliar with who gathered Wi-Fi or residence hall access logs and did not know how the digital information had been compiled. When The College Voice asked Rothenberger if she knew who was leading the data-gathering efforts, Rothenberger said “I don’t know the answer to that question. The data came to me… I didn’t talk with the person who ran the report,” referring to the querying of College databases containing Wi-Fi or residence hall access logs. Rothenberger also claimed that she was not involved in the decision-making process regarding what database time frames to look at with respect to Wi-Fi or residence hall access information: “who originally decided what times the report will be run on is someone else. Who decides who should be interviewed off of that report? That was us.”
Manning suggested that Arcelus was the one who “ordered” data from Information Services and indicated that she did not play a role: “I want to say that Victor [Arcelus] ordered, worked with I.T., because that information was sent to me and was a certain list of people. So I’m not sure how they narrowed it down.” Manning noted that, with respect to data-gathering tactics used in the investigation, she is “not well versed in… forensic digital stuff.”
Manning denied requests from The College Voice to share documents relating to the preliminary inquiry or the College’s investigation.
In a January 28 interview, after the College had begun sharing data with police, Manning suggested that the Title IX office is not equipped to conduct digital-focused investigations: “What the school has to understand, what students have to understand is that Title IX, we have, you know, we don’t have forensic capabilities — like, I can’t do a deep dive into things as the New London London Police Department can, like, they have the forensic capabilities to go into computers and this and that. What we’ve done is look at, and we’ve said this, we’ve looked at who’s gone in and out of the building, whose phone has picked up on our wireless network around the times that’s happened and so that’s what we’ve done. We’ve handed all of that over to the police.”
In an interview following Alberti’s arrest, Arcelus suggested that querying residence hall access log records is not a complex task.
“The door access points is a really easy kind of spreadsheet that can be run very simply. My understanding is that mining the information out of the system to get to the wireless access points is a more complex function which I think took more time to be able to to pull together.”
According to Savage, residence hall access logs are stored in a database that is not maintained by campus safety, but that “campus safety can access if they need immediate info on a particular student’s access records.” Savage said that if “more expansive information is needed, campus safety asks [Information Services] to provide a report with the needed information. Lee Hisle, the College’s Vice President for Information Services, told The College Voice in a May 9 email that he wouldn’t comment on an “ongoing investigation.”
It remains unclear to what extent Rothenberger used residence hall access logs, if at all, during her inquiry, despite her apparent use of Wi-Fi access logs. Prior to Alberti’s identification and arrest, Manning told The College Voice that Rothenberger was using both W-Fi and residence hall access logs in the investigation. In recent exchanges, Savage has said that these records were used in the internal investigation, and Arcelus said he was under the impression that investigators were using these logs.
But, in a March 28 email, Manning told The College Voice that initial datasets used in Rothenberger’s investigation did not include residence hall access logs: “initially, during the preliminary inquiry, we focused on which phones picked up the Wi-Fi in Plant and Morrison; we did not look at key card access at that time.” Manning later said that she could not clarify whether residence hall access logs were ever included in the investigation before the College brought in external investigators from Ankura, a consulting firm, in January after the fifth incident.
“Their list was way longer than ours”
The residence hall access logs that police say placed Alberti at the scene of five reported incidents between October and January do not appear to have been used in the College’s investigation until the College transferred investigative responsibilities from Rothenberger to Ankura in January after the fifth reported incident. But in the two weeks following the conclusion of Rothenberger’s inquiry, both Ankura and NLPD identified Alberti as a suspect in the case.
According to Manning, Ankura “took a deeper dive” into a “more extensive” set of Wi-Fi and residence hall access logs. Manning said Ankura’s investigators “asked questions that we probably didn’t know to ask, or some person, the people who [originally] sent us the data, maybe didn’t know to ask.” Manning said that she compared the list of possible witnesses that she had provided to Rothenberger with a list of names that Ankura came up with: “their list was way longer than ours.” Soon after the College brought in Ankura, the investigators had narrowed down a list of possible suspects to “two people…including [Alberti] and another person,” Manning claimed.
According to Manning, Ankura sent Alberti’s name to NLPD “along with all of the door access and Wi-Fi access.” Chief Reichard refused to provide the dates on which NLPD investigators learned about or recieved access to residence hall access logs, but Savage has said that the College first shared the data in early January:
“The college began communicating with NLPD in December [2018]. After consulting with college counsel on what we could and could not share legally we began providing them with information and documentation on January 2 [2019].”
Savage clarified that “regulations including FERPA and Title IX” required legal consideration.
Arcelus also indicated that NLPD still had not received full access to such records until January: “In January, there was a meeting with NLPD and our [Information Services] department in order to be able to make sure that we were able to get to them exactly what they needed so that everybody was on the same page about getting, again, the right kind of data together.”
NLPD became involved in the case after the first voyeurism incident on October 17. Police say they received complaints regarding that first incident, as well as an October 25 incident in Morrison and a December 5 incident in Plant.
But an affidavit requesting a warrant for Alberti’s arrest that details the timeline of the NLPD investigation also suggests that police were not officially aware of the November 30 incident in Plant until late January. Only on January 25, two days after a fifth reported incident in which the same student who reported the November 30 incident reported another experience of being recorded in a Plant shower, did Savage inform police about these two incidents.
The police affidavit highlights an email sent from a victim to Savage and Manning on December 6, 2018, at 7:42 PM informing the two that “Around 12:30a.m. last Thursday [11/30/18] I saw what could have been a camera directed at me while I was in the shower.” The email was sent just two hours after Savage and Manning had sent a campus-wide email informing students of the December 5th Plant incident. Earlier that afternoon, NLPD was investigating that case in Plant, according to the email from Manning and Savage. NLPD say that the victim who reported the November 30 incident did not want to talk to police.
“If we’re able to catch somebody in the act of doing it, then that starts to change things”
In the same December open forum in which Arcelus mentioned the College’s ability to use residence hall access logs for investigations, Arcelus suggested that it would be very difficult to solve the voyeurism investigation without community participation: “these kinds of investigations have all kinds of dead ends and they are very difficult to do because there is typically not a lot of information available. If we’re able to catch somebody in the act of doing it, then that starts to change things, then that enables us to be able to engage that individual and engage NLPD in ways that can be particularly productive in getting to the bottom of it.”
In interviews following Alberti’s arrest, both Manning and Arcelus suggested that student involvement played a role in identifying Alberti as a suspect. Manning noted that a student tip brought Alberti’s name to her attention for the first time on January 28: “we actually had a student come in and give us a tip and said that this particular person was acting strangely in a bathroom.” The College passed on that information to Ankura investigators, and Manning said Ankura matched the name to records in their possession: “they looked at the door access or the keycard access and the Wi-Fi and all that and then said ‘I think this is the person.’”
The College Voice spoke with Hans Horst-Martz ‘21, who says he contacted Dean of Institutional Equity and Inclusion John McKnight after a bizarre bathroom interaction with Alberti. Horst-Martz said that on January 26, 2019, around 5:30 PM, he was in his Wright House bathroom washing dishes: “I saw the door open about half-a-foot and it was kind of just held there. I could see the edge of a hood, but I couldn’t see a face. Right before I would be able to see their forehead, they left.
“There in the hallway is Carlos [Alberti] and I was like ‘Carlos, are you lost? Are you looking for someone?’ and he was like, ‘oh I’m really high right now, I’m trying to find Harris.’”
Horst-Martz said this was especially odd, because this occurred a residential space in immediate proximity to Harris dining hall: “I live on the second floor of Wright and it’s so easy to get to Harris, you literally just walk out the door and you’re there.” Minutes later, Horst-Martz says Alberti texted him saying that he found Harris.
Horst-Martz says that soon after he received the text from Alberti, he went to Harris for dinner. But, Horst-Martz said that later, when he was leaving the dining hall, “I saw [Alberti] walking in alone, as if he’s coming to dinner — like 45 minutes after he texted me that he already found Harris.” The next day, Horst-Martz emailed McKnight asking to meet about the interaction. Horst-Martz met with McKnight and Manning on January 28 at which point he says he first told them about Alberti.
Manning described Horst-Martz’s report as “the first time [Alberti’s] name was on my radar.” Arcelus said that “student tips played a significant role in being able, as far as I understand … to narrow that down.”
But it remains unclear whether a student tip such as Horst-Martz’s was necessary for internal College investigators, Ankura investigators, or NLPD to analyze datasets like the residence hall access logs that NLPD and Manning have said placed Alberti at the scene of five reported incidents. Chief Reichard has told The College Voice that the College’s internal investigation did not play a role in how NLPD identified Alberti as a suspect.
Arcelus says he is confident in the College’s investigative process, but insisted that student involvement is a requirement: “It might not require student involvement on a grand scale of everybody on campus, doing what we were asking people to do during that period of time. But every investigation requires input from students, interviews of the people involved, interviews of witnesses. It’s very dependent on students. So I think that … has been and will continue to be an integral part of whatever work we do on any kind of investigations in the future.”
Horst-Martz took issue with the idea that students should shoulder that responsibility: “It’s not my job to play security guard for my dorm.”
“The reality is that an incident like this — it’s like finding a needle in a haystack”
In a January 28 interview, Manning suggested that the datasets that informed the preliminary inquiry may not have been sufficient because of the possibility of students evading Wi-Fi or door access detection: “Are we missing people? Were there other people in the building who just turned [their] Wi-Fi off until we just didn’t pick them up? Are there people in the building who didn’t swipe and they were being walked in with someone? So it’s not — it’s the system … I mean it’s a good system. But there are pieces that are missing.”
In an interview following Alberti’s arrest, Arcelus suggested that the College could have responded quicker if it had surveillance cameras at the entrances of residence halls, saying that cameras would have provided another type of “technological information” that may have helped track down the suspect. “It’s a reasonable step for the college to take, to place cameras at entrances and exits of buildings which would then give us additional information about entry into buildings, different than card access. Because with card access, somebody can let you in and you’re not identifiable as having come into the building. But, with cameras, then that more readily collects information about people coming and going.”
“The reality is is that an incident like this — it’s like finding a needle in a haystack,” said Arcelus. “We have to be able to try to bring to bear whatever technological information, technologically-based information we have, plus with whatever tips we have, to help connect the dots enough to be able to narrow a campus of X number of students, faculty, staff, down to a reasonable number of people that might be responsible.”
Arcelus clarified that he was “not suggesting having cameras in bathrooms or hallways.”
In a May 10 email to the campus community, college president Katherine Bergeron provided a list describing measures the College is “prepared to do” in the coming months. Listed first is “phased installation of cameras to increase security at entrances and exists to residence halls.”
Although video logs of entrances might have provided the College additional information in the case, NLPD and Manning have both said Alberti’s name was found in datasets that the College already had. Alberti does not appear to have avoided card readers at the entrances to Plant and Morrison, because police have said their analysis found his name in logs, and Manning has said Ankura had similar results.
It is also unclear whether the College is any better equipped to conduct an investigation using unlabeled surveillance footage than it was to use and compare existing Wi-Fi and residence hall access logs that already contain structured identifiable information, including, in this case, the culprit’s name.
“It was a compulsion”
The February affidavit requesting a warrant for Alberti’s arrest detailed Alberti’s alleged history of recording women in campus co-ed bathrooms and on New York City subways, expanding the scope of “shower incidents” from five that were reported to a set of 213 different shower stall videos that included approximately 30 women in eight different buildings at the College.
72 videos were also found on Alberti’s electronic devices in which he allegedly recorded women on New York City subways from beneath their skirts, allegedly showing women’s “genitals, pubic area, or buttock or the undergarments that clothe the genitals, pubic area, or buttock of a woman.” According to NLPD, Alberti “needed to know what was under their skirt.” He “felt it was a compulsion and could not stop.”
Alberti admitted that early on, he would walk into dorm bathrooms hoping to run across a woman showering. NLPD says Alberti eventually took to sitting in campus bathrooms “for extensive periods of time, sometimes up to three hours, waiting for a woman to use the shower so he could record them.”
Alberti has been expelled from the College, according to a college-wide email from Bergeron. He remains released on $150,000 cash bond and is next scheduled to appear in New London Superior Court on May 30.
Saadya Chevan contributed reporting.
