DISSENTING OPINION BY
Judge COHN JUBELIRER.Because the Workers’ Compensation Judge (WCJ) found that Claimant suffered a psychic workplace injury in the nature of Post Traumatic Stress Disorder (PTSD), the occurrence and causation of which are incontrovertibly attributable to a specific abnormal workplace event (Event),1 I would conclude that this workplace injury is compensable under the Pennsylvania Workers’ Compensation Act (Act).2 Because I would affirm the Workers’ Compensation Appeal Board (Board), I respectfully dissent.
As our Supreme Court has concluded, a claimant is entitled to workers’ compensation (WC) benefits for a psychic injury if he can prove by objective evidence that he has suffered such an injury and that the injury is more than a subjective reaction to normal working conditions. Martin v. Ketchum, Inc., 523 Pa. 509, 519, 568 A.2d 159, 164-65 (1990). In the instant case, the WCJ credited the testimony of Claimant, his medical witness, Dr. Raditz, and Employer’s medical witness, Dr. Michals, all of which established that Claimant suffered PTSD and attributed that injury to the work-related Event. (WCJ Decision, Findings of Fact (FOF) ¶¶10, 11, 12.) Thus, both Employer and Claimant have established through the credited testimony of their witnesses that Claimant suffered work-related PTSD and that the work*112place Event indisputably caused this injury.
In RAG (Cyprus) Emerald Resources, L.P. v. Workers’ Compensation Appeal Board (Hopton), 590 Pa. 413, 912 A.2d 1278 (2007), our Supreme Court set forth the following' two-prong analysis for its review in cases involving a WCJ’s grant of WC benefits for a psychic injury: (1) “whether the Commonwealth Court abused its discretion by substituting its factual findings for those made by the WCJ and supported by the record”; and (2) “whether the findings of fact support the legal conclusion that Claimant’s injury was the result of an abnormal working condition.” RAG (Cyprus) 590 Pa. at 426, 912 A.2d at 1286.3 Determining whether a particular workplace scenario causing injury to a claimant is “normal” or “abnormal” is a highly fact-sensitive inquiry, requiring “deference to the factual findings of the WCJ,” who had the benefit of observing the witnesses. Id. at 425, 912 A.2d at 1286.
I respectfully believe that the majority opinion would not pass this two-prong analysis in the instant case. The majority opinion reverses the grant of WC benefits here, without first addressing whether the WCJ’s factual findings are supported by the record, and prematurely reaches its own, different conclusion that the Event was a normal working condition and, therefore, not compensable. This approach is problematic for two reasons. First, the majority opinion does not limit its review to the WCJ’s factual findings and whether those factual findings are supported in the record, but instead relies upon uncredited evidence to essentially make its own factual findings that support its conclusion. Second, the majority opinion overemphasizes the role of the “foreseeability” in determining whether a workplace event is normal or abnormal, without performing the required, highly fact-specific inquiry into the Event.
I will address each of these concerns in turn, but preliminarily note that this Court’s scope of review is limited to whether there is substantial evidence to support the WCJ’s findings of fact. Bethenergy Mines, Inc. v. Workmen’s Compen*113sation Appeal Board (Skirpan), 531 Pa. 287, 291, 612 A.2d 434, 436 (1992). Substantial evidence is that relevant evidence a reasonable person “might accept as adequate to support a finding.” York Terrace/Beverly Enterprises v. Workmen’s Compensation Appeal Board (Lucas), 140 Pa.Cmwlth.75, 591 A.2d 762, 764 n.5 (1991). Additionally, we should take all inferences drawn from the evidence in favor of the party prevailing before the WCJ. Krumins Roofing and Siding v. Workmen’s Compensation Appeal Board (Libby and State Workmen’s Insurance Fund), 133 Pa.Cmwlth.211, 575 A.2d 656, 659 (1990). Finally, the WCJ is the ultimate determiner of credibility. Greenwich Collieries v. Workmen’s Compensation Appeal Board (Buck), 664 A.2d 703, 706 (Pa.Cmwlth. 1995). It is the WCJ’s function “to weigh the evidence and resolve conflicting testimony.” Alpo Petfoods, Inc. v. Workmen’s Compensation Appeal Board (Neff), 663 A.2d 293, 295 (Pa.Cmwlth.1995). The WCJ is free to accept, in whole or in part, the testimony of any witness. Greenwich Collieries, 664 A.2d at 706.
My first concern is that the majority opinion does not limit its review to the WCJ’s factual findings and whether those factual findings are supported in the record. Here, the WCJ found Claimant’s testimony credible, emphasizing her personal observation of his manner and demeanor during testimony and the consistency of his testimony and medical evidence. (WCJ Decision, (FOF) ¶ 10). The WCJ rejected Employer’s evidence to the extent to which it was not consistent with Claimant’s testimony. (FOF IT 11.) The WCJ, who had the benefit of viewing the witnesses, was well within her discretion to rely upon the testimony of Claimant and to reject that of Employer’s witnesses. However, the majority does not begin its analysis by examining whether the WCJ’s factual findings are supported by credited, substantial evidence in the record. Instead, the majority focuses on evidence in the record, not all of which was credited by the WCJ, such as Employer’s training materials, pamphlets, and statistics on robberies of Employer’s stores in southeastern Pennsylvania, and makes its own factual findings reaching a different conclusion than that reached by the WCJ. For example, the majority refers to evidence that Employer provided Claimant with pamphlets and educational tools on how to handle a robbery in addition to training on workplace violence, some of which addressed robberies and thefts. PA Liquor Control Board v. Workers’ Compensation Appeal Board (Kochanowicz), 29 A.3d 105, 110-11 (Pa.Cmwlth.2011). The majority concludes that, because Claimant admitted that he attended Employer’s trainings and received the educational materials, “Claimant could have anticipated being robbed at gunpoint.” Id. However, one of Employer’s pamphlets relied upon by the majority states, “Generally, robberies occur very infrequently.” (Pamphlet, R.R. at 119a (emphasis added).) This pamphlet, which describes robberies as “infrequent” and thus not normal, contradicts Employer’s and the majority’s position and supports the findings of the WCJ. Moreover, the majority relies on the mere existence of these materials, along with two or three training sessions in a span of Claimant’s thirty-year career, to transform the Event into a normal working condition. This is contrary to the WCJ’s finding that such materials were provided to help employees know how to behave in those “infrequent” situations to “best ensure[ ] the safety of the person to whom the gun is pointed, as well as fellow employees and customers” if ever confronted with a workplace crime. (FOF ¶ 13.) *114The WCJ, therefore, “[did] not find [the training materials] entirely relevant” to the Event Claimant experienced in this case. (FOF ¶ 13.)
Similarly, the majority relies on Employer’s statistics, which the WCJ also did not credit, (FOF ¶ 11), to hold that “robberies of liquor stores are a normal condition of retail liquor store employment in today’s society.” PA Liquor Control Board, 29 A.3d at 111. Those statistics purported to reveal that there had been ninety-nine armed robberies in its southeastern Pennsylvania retail stores since 2002. Id. However, there is no explanation as to why Employer uses its statistics from southeastern Pennsylvania, as opposed to Bucks County (where Claimant’s store was located), or why Employer did not include a larger area or even the entire state in the statistical evidence. Moreover, there is no methodology explaining how or why the proffered statistics, dating from 2002 for a five-county area, are relevant to prove that this Event is not an abnormal working condition. The statistics do not provide any meaningful insight about the frequency of the same type of event that occurred here. In fact, those statistics are potentially misleading given the number of liquor stores in that five-county area and the length of time the statistics cover. Indeed, in Bucks County, there were only three robberies referenced in the statistics, one of which was the Event, one that occurred only within a few weeks of the Event, and one that occurred after the Event. (Workplace Violence Report, February 12, 2009, R.R. at 330a; Workplace Violence Report, April 28, 2008, R.R. at 334a; Workplace Violence Report, March 1, 2008, R.R. at 339a.) Moreover, because there is no talismanic number that transforms an abnormal work incident into a normal one, this Court must look to the totality of the circumstances. RAG (Cyprus), 590 Pa. at 430 n. 10, 912 A.2d at 1289 n. 10. Thus, I believe the WCJ was justified in not crediting the statistics presented. Accordingly, I believe the majority’s reliance upon evidence the WCJ did not credit in order to support its conclusion is beyond our appellate scope of review as set forth in RAG (Cyprus).
Confining my review to the evidence credited by the WCJ, and examining the record, including all inferences reasonably drawn therefrom in the light most favorable to Claimant, I conclude that there is substantial evidence in the record to support the WCJ’s findings. Claimant’s credited testimony establishes that the Event was not a normal work condition. Claimant credibly testified that he had worked in or managed over twenty of Employer’s l'etail stores for more than thirty years, and he testified that he had never been involved in anything similar to the Event or any kind of robbery. (WCJ Hr’g Tr. at 12, R.R. at 37a.) If the Event was a normal working condition, the WCJ could conclude that, in over thirty years of experience in more than twenty of Employer’s stores, Claimant would have encountered such events. In her findings of fact, the WCJ specifically noted that Claimant was not aware of any specific details of robberies and did not track violence in the area of his store. (FOF ¶ 9.) The abnormality of the Event, particularly in Claimant’s store, is further highlighted by the WCJ’s finding that Employer had hired security guards at some stores, but did not provide a security guard for Claimant’s store. (FOF ¶ 9.) Although Claimant acknowledged that the clientele of this store and others he previously managed raised shoplifting concerns, (WCJ Hr’g Tr. at 20, R.R. at 45a), shoplifting is entirely different from what occurred during the Event. Importantly, Claimant’s testimony that, unlike other stores, Employer did not provide Claimant’s store with an alarm system *115or emergency alarm button, (WCJ Hr’g Tr. at 23, R.R. at 48a), offers additional support to the WCJ’s finding of an abnormal working condition in this matter. Employer’s decision not to install alarm systems or have guards in Claimant’s store undermines the Employer’s contention, with which the majority agrees, that armed robberies are normal occurrences since Employer apparently did not consider them likely enough to occur to protect against them. Even the trainings on handling unlikely workplace events were infrequent, with Claimant attending only two or three in his thirty-plus years with Employer. There also is no evidence that Employer provided any protective equipment or gear, as is provided to employees whose jobs entail great personal safety risks as a normal part of their work, such as police or corrections officers. (FOF ¶ 9.) The WCJ could conclude that the armed robbery here, which included having a masked man point a gun at Claimant, bind and tie him to a chair, and prod his head with a gun while inquiring whether Claimant was impatient with him, was an intense event and beyond a standard robbery. I believe that such credited evidence constitutes substantial evidence on which the WCJ could base her finding that the Event was an abnormal work condition.
The majority opinion cites McLaurin v. Workers’ Compensation Appeal Board (SEPTA), 980 A.2d 186 (Pa.Cmwlth.2009), and Kennelty v. Workers’ Compensation Appeal Board (Schwan’s Home Service, Inc.), 594 Pa. 12, 13, 934 A.2d 692, 692 (2007) (per curiam), to support its conclusion that the frequency of the occurrence of a particular work condition is considered in determining whether it is an abnormal occurrence. In McLaurin, a SEPTA bus driver of six months suffered from PTSD and related conditions after an incident in which several hooded young men entered his bus without paying their fares and, just before disembarking, one of the young men pulled out a gun and caused the driver to believe he was going to be shot. McLaurin is distinguishable because, in that case, the WCJ credited the employer’s witnesses as to the frequency of operator assaults, the training provided and, based upon that credited evidence, determined that the incident was not abnormal. McLaurin, 980 A.2d at 189. The case at bar is more similar to Kennelty, in which the Supreme Court reversed this Court’s determination, holding that this Court disturbed the WCJ’s credibility determination. In Kennelty, the WCJ had found the employer’s evidence credible that the frequency of work-related incidents “experienced by [the claimant] was normal for [the employer’s] specific industry” and directed that this Court is “not free to disturb this credibility determination based on competent evidence.” Kennelty, 594 Pa. at 13, 934 A.2d at 692. The Supreme Court emphasized that the WCJ is the ultimate finder of fact and the exclusive arbiter of credibility and evidentiary weight. Id. Here, as in Kennelty, the WCJ found a party’s evidence credible and gave more weight to that evidence; however, notwithstanding those findings, the majority seeks to disturb that credibility determination by relying on other evidence to reach a conclusion contrary to the WCJ’s determination. I believe that such disturbance exceeds our scope of review and conflicts with the Supreme Court’s admonition in Kennelty.
Next, I believe that the majority opinion overemphasizes the role played by the foreseeability of any given workplace event to transform it into a normal working condition. In this regard, the majority opinion seemingly equates “foreseeability” with “normalcy.” However, the fact that nearly anything is foreseeable, including a robbery, because robberies do occur, does not *116make that event “normal.” Although I agree with the majority’s statement that there is no bright line test or general standard that we consider in determining whether a particular event is normal or abnormal, I am unwilling to accept the premise that simply because robberies are known to occur, they are a “normal” condition of the workplace. Moreover, because not all robberies are identical, we should not treat them categorically as if they were. It is well settled that these matters require a highly fact-sensitive inquiry into exactly what occurred on a case-by-case basis. Payes v. Workers’ Compensation Appeal Board (Commonwealth of PA/State Police), 5 A.3d 855, 859-860 (Pa.Cmwlth.2010), appeal granted, — Pa. -, 20 A.3d 1182 (2011).4 Indeed, in RAG (Cyprus), the WCJ considered the type of “jovial antics” at issue in that case, recognizing that some were normal, but that the offending comments in that case were not normal, and our Supreme Court affirmed. RAG (Cyprus), 590 Pa. at 421, 912 A.2d at 1283. Were we to paint all robberies with the same brush and take the position that robberies are foreseeable and, therefore, an armed robbery is normal, we would abdicate our responsibility to review these matters on a case-by-case basis. Such a broad, inflexible position is not appropriate in matters where this Court is required to perform a highly fact-specific inquiry.
Finally, I question whether the “abnormal working condition” standard should be applied in this matter at all. The requirement that a claimant prove that a psychic injury was not just a subjective reaction to normal workplace events arose from the “inherent difficulty in establishing causation ... because such maladies are intrinsically subjective.” RAG (Cyprus), 590 Pa. at 428, 912 A.2d at 1287. In Martin, the Supreme Court reviewed the history of psychic injury cases and agreed that the abnormal working condition analysis “was intended to distinguish psychiatric injuries that are compensable because the necessary causal relationship between the employment and mental disability has been established from those psychiatric injuries that arise from the employee’s subjective reactions to normal working conditions.” Martin, 523 Pa. at 518-519, 568 A.2d at 164 (emphasis added). Thus, Martin’s focus was on whether the psychic injury was, in fact, work related, or whether it was actually the result of additional non-work factors, such as personality and individual vulnerability. However, unlike in other cases, such as RAG (Cyprus) and Martin, the occurrence of Claimant’s psychic injury is undisputed and its cause has been found to be the Event by both Claimant’s and Employer’s expert witnesses. Thus, the concern about causation present in psychic injury cases, and which the “abnormal working condition” test addresses, is not present here. Therefore, I question whether that test is applicable under these circumstances.
Accordingly, I would affirm the Order of the Board.
Judge McGINLEY and Judge BUTLER join this dissenting opinion.
. The Event took place on April 28, 2008 at the Bucks County retail liquor store Claimant managed. The Event consisted of an armed robbery during which a masked gunman directed Claimant’s movements and actions; tied Claimant up; bound him to a chair in a back room, whereupon Claimant lost his freedom of movement; kept a gun pointed at Claimant; and, while expressing impatience and frustration with Claimant, prodded the gun against Claimant’s head. (WCJ Decision, Finding of Fact ¶ 2.)
. Act of June 2, 1915, P.L. 736, as amended, 77 P.S. §§ 1-1041.4, 2501-2708.
. In RAG (Cyprus), a case involving an aggravation of a mineworker’s pre-existing PTSD caused by a supervisor's sexually harassing comments, the WCJ made findings that such comments were not normal occurrences in the mines although a “lesser level of jovial antics was common.” RAG (Cyprus), 590 Pa. at 421, 912 A.2d at 1283. Based upon the findings of fact, the WCJ concluded that the supervisor’s comments constituted abnormal working conditions, emphasizing the calculated nature and intensity of the comments from the normal joking or uncivil behavior that often took place in the mines. Id. at 421-422, 912 A.2d at 1283-1284. The Board affirmed the WCJ, noting that several co-workers had testified that the comments went beyond those accepted in the mines and agreed, "[biased on the WCJ's findings of fact ... the claimant had established the existence of abnormal working conditions.” Id. at 422, 912 A.2d at 1284. However, this Court reversed because it determined that the "evidence fail[ed] to support a finding of abnormal working conditions.” Id. On appeal, the claimant asserted that this Court had usurped the function of the fact-finder by reweighing the evidence. In its review, the Supreme Court applied the two-prong examination set out above. With regard to the first prong, the Supreme Court concluded that this Court abused its discretion by not limiting its review to whether the WCJ's factual findings were supported by the record, noting three instances in which this Court essentially made its own interpretations of the record, and by focusing on a section of testimony not included in the WCJ’s factual findings to support this Court’s own conclusion. Id. at 426-427, 912 A.2d at 1286-1287. On the second prong, the Supreme Court reviewed whether the facts as found by the WCJ established the claimant’s right to compensation pursuant to Martin, and held that they did. Id. at 429, 912 A.2d at 1288.
. The question on appeal to the Supreme Court, as framed by the petitioner in that case, is whether this Court "erred as a matter of law in concluding that the claimant was not exposed to abnormal working conditions when the WCJ found that he was exposed to an unusual [work] event which made his job more stressful than it had been.” Payes, — Pa. at -, 20 A.3d at 1182.