*166{¶ 1} In this appeal, the state challenges a decision of the court of appeals granting the petition for postconviction relief of the defendant-appellee, Willie Herring, a death-row inmate. The state challenges the appellate court’s finding that Herring’s counsel provided ineffective assistance in failing to properly prepare for the mitigation phase of trial and the court’s decision to order a new sentencing hearing.
{¶2} We hold that the court of appeals properly applied the standards set forth in Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984), in holding that trial counsel were deficient in preparing for mitigation and that the deficiency was prejudicial. Accordingly, we uphold the decision of the court of appeals vacating the death penalty and remanding the matter to the trial court for a new sentencing hearing.
I. Facts
{¶ 3} Evidence presented at trial showed that Herring and five other individuals robbed the Newport Inn, a bar in Youngstown, shortly after midnight on April 30, 1996. They shot five people, robbed the till, and left. Three of the five victims died and two others were seriously wounded. See State v. Herring, 94 Ohio St.3d 246, 762 N.E.2d 940 (2002).
{¶ 4} Herring was the evident ringleader of these crimes. Id. at 266. The participants had gathered at Herring’s house before the robbery. Id. at 246. Herring provided three of the individuals with handguns and kept a 9 mm Cobray semiautomatic for himself. Id. at 246-247. Herring donned a Halloween mask, which was a store-bought mask similar to one seen in “slasher” movies. No one else had a similar mask; the others hid their faces with bandanas or a T-shirt. Id. at 247.
{¶ 5} Ronald Marinelli, the Newport Inn’s owner, was tending bar that night. Several customers were in the bar, including Deborah Aziz, Herman Naze Sr., Dennis Kotheimer, and Jimmie Lee Jones. Id.
{¶ 6} Upon bursting into the bar, one of the gunmen ordered Naze to “[g]ive me your fucking money.” Id. When Naze stated that he did not have any money, the gunman shot him. Then Herring shot Aziz, who fell to the floor. She managed to crawl away and hide behind a cooler and a trash can. She later described her assailant’s mask as “a hard plastic, like one of those Jason masks.” Id.
{¶ 7} Herring then walked around the end of the bar and approached Marinelli and the cash register. He shot Marinelli four times in the stomach from about five feet away. Marinelli managed to stay on his feet as Herring came closer. Id. at 247. Herring told Marinelli, “Give me your fucking money.” Despite his *167wounds, Marinelli obeyed, handing over the cash in the register. Id. at 248. But Herring screamed that Marinelli had not given him everything. As Herring threatened to “blow [Marinelli’s] brains out,” Marinelli gave him money from a nearby drawer. Herring then screamed for more money. Marinelli urged him to “[b]e cool” and told him that there was no more. Herring responded by leveling his gun at Marinelli’s head. Id.
{¶ 8} Marinelli reached into the drawer again and pulled out a gun. But Marinelli was so weak that Herring easily took it away from him. Marinelli collapsed. Herring then said, “You ain’t dead yet, motherfucker,” and shot Marinelli in the legs as he lay on the floor. Id.
{¶ 9} After Herring shot Marinelli, Aziz heard Kotheimer say, “You motherfucker.” Then she heard more shots. Marinelli saw Kotheimer get shot but did not see who shot him. Nobody saw who shot Jones. Id.
{¶ 10} Following the arrival of the police, the shooting victims were taken to a Youngstown hospital. Naze, Jones, and Kotheimer died from their wounds. Id.
II. Trial results and the mitigation hearing
{¶ 11} At trial, Herring was convicted of three counts of complicity to commit aggravated murder, two counts of attempted aggravated murder, two counts of aggravated robbery, and firearm specifications. 94 Ohio St.3d at 248-249, 762 N.E.2d 940. As to Count One, the murder of Jones, the jury found Herring not guilty as the principal offender but guilty as an accomplice. Herring was also found guilty of three course-of-conduct death-penalty specifications that were attached to each of the murder counts, pursuant to R.C. 2929.04(A)(5). Id. at 249.
{¶ 12} During the mitigation hearing, trial counsel presented evidence showing that Herring’s accomplices did not receive the death penalty. The defense also presented the testimony of Deborah Herring, the defendant’s mother, and Nicole Herring, the defendant’s sister.
{¶ 13} Deborah testified that Herring has six brothers and sisters. She testified that Herring had a good and loving relationship with his siblings and continues to stay in touch with them. Herring helped with chores and helped take care of his younger brothers and sisters when he was growing up. Deborah asked the jurors to spare her son’s life.
{¶ 14} Nicole testified that Herring helped to take care of his younger brothers and sisters. She stated that she had a close relationship with Herring and that they did numerous things together. Finally, Nicole asked the jury to spare her brother’s life.
{¶ 15} On rebuttal, Timothy Franken, an assistant prosecuting attorney, testified that Antwan Jones, one of the accomplices, was originally charged exactly *168the same as Herring. The death-penalty specifications, however, were voluntarily dismissed against Jones because the prosecutors did not think that they could prove them. Franken also testified that Adelbert Callahan was charged exactly the same as Herring, but Callahan could not receive the death penalty because he was a juvenile at the time of the crimes. Franken also mentioned that Eugene Foose, another accomplice, was a juvenile. The two other accomplices had had lesser culpability. Louis Allen did not shoot anyone; indeed, he ran away as soon as the shooting started. Kitwan Dalton, the getaway driver, neither entered the Newport Inn nor carried a weapon. 94 Ohio St.3d at 268, 762 N.E.2d 940.
{¶ 16} During closing arguments, trial counsel emphasized that Herring had been convicted as an aider or abettor and that he had been found not guilty as a principal offender in committing the murders. Trial counsel also argued regarding the disparity in sentencing that would exist if Herring were sentenced to death when the state had not pursued death sentences for his accomplices. In addition, counsel raised Herring’s youth as a mitigating factor; he was 18 at the time of the offenses.
{¶ 17} After the mitigation hearing, the jury recommended death for all three aggravated murders. The trial court sentenced Herring to death. Id. at 249.
{¶ 18} On February 27, 2002, this court affirmed Herring’s convictions and death sentence. Id. at 269. The United States Supreme Court denied certiorari. Herring v. Ohio, 537 U.S. 917, 123 S.Ct. 301, 154 L.Ed.2d 202 (2002).
III. Postconviction proceedings
A. Herring’s postconviction claim
{¶ 19} On September 17, 1999, Herring filed his petition for postconviction relief. He requested discovery and an evidentiary hearing. Herring’s primary claim challenged trial counsel’s effectiveness in faffing to conduct an adequate mitigation investigation prior to his mitigation hearing. Herring presented affidavits and other documentary evidence in support of his claim.
1. Affidavits from experts and family members
{¶ 20} Dr. Jolie Brams, a psychologist, submitted an affidavit stating that the jurors were presented “no meaningful psychological information” about Herring’s “problematic functioning over the course of his life.” She stated that no lay or expert testimony was presented that “would have given the jurors an opportunity to understand the sociocultural, psychological, developmental and intellectual factors that operated to bring Mr. Herring to that particular point in time.”
*169{¶ 21} Dr. Brams indicated that the testimony of Herring’s mother and sister during the mitigation hearing presented a distorted picture of Herring’s upbringing. As to their testimony, Dr. Brams stated: “The jurors only saw two supposedly caring individuals which did nothing to educate them regarding the marked dysfunction in Mr. Herring’s family and the amazingly dysfunctional role models by whom this young [man] was raised. Instead, they were presented a picture of a family that cared.”
{¶ 22} Dr. Brams discussed her own evaluation of Herring’s family, his history, and his past and present functioning that could have been presented during mitigation. She stated that Herring’s “childhood was remarkably dysfunctional in almost every aspect.” She stated that Herring “was raised in an environment in which basically every parental figure, caregiver, family member, and associate was involved in illegal activities, significant drug and alcohol abuse, and to whom the consequences of violating the law evoked little anxiety.” Dr. Brams described Herring as a “ ‘feral child,’ who roamed the neighborhood aimlessly, without any adult having meaningful or consistent concern.”
{¶ 23} Dr. Brams stated that Herring began using marijuana and alcohol on a frequent basis by the time he was 13 or 14. Herring’s active substance abuse was a primary factor leading to his failure in school and his eventual decision to drop out of school.
{¶ 24} Herring’s family members introduced him to gangs and drug dealing, which was described to Dr. Brams as “the family business.” Dr. Brams added that Herring’s mother asked him to purchase drugs for her own consumption and that his mother and other family members used drugs openly.
{¶ 25} Dr. Brams stated that no psychometric tests were conducted prior to trial assessing any aspect of Herring’s mental capacities. There was also no exploration of his intellectual and academic capacity from other sources, such as interviews with Department of Youth Services (“DYS”) personnel and public school teachers or a review of his institutional records. Dr. Brams also stated that a neuropsychological evaluation should have been conducted prior to trial to establish whether Herring suffered from organic brain impairment.
{¶ 26} Dr. Brams’s staff administered various tests to Herring. Test results on the Wechsler Adult Intelligence Scale-Third Edition (“WAIS-III”) showed that Herring had a verbal IQ of 85, a performance IQ of 91, and a full-scale IQ of 87. Results on the Beery Buktenica Developmental Test of Visual-Motor Integration placed Herring in the third percentile nationally and showed his difficulty with visual-motor functioning. Results on the Matrix Analogies Test showed that Herring has the perceptual-learning skills of a ten-year-old.
{¶ 27} Dr. Brams also indicated that Herring’s DYS records provided information about Herring’s positive characteristics. She stated that the DYS records *170“reflect that Mr. Herring was able to meet the expectations of the staff in many areas and complete goals that were set out for him in terms of socialization and education.”
{¶ 28} Finally, Dr. Brams set forth her diagnoses of a number of psychiatric disorders as defined by the Diagnostic and Statistical Manual of Psychiatric Disorders. She concluded that Herring presents (1) alcohol abuse, chronic, (2) cannabis abuse/dependence, (3) polysubstance abuse/dependence, (4) depressive disorder, (5) personality disorder with narcissistic and antisocial features, and (6) learning disabilities.
{¶ 29} Dr. C. Ronald Huff, the director of the School of Public Policy and Management at Ohio State University, submitted an affidavit about Herring’s gang involvement. Dr. Huff stated that Herring “grew up with a host of gang members as role models, especially on his father’s side of the family.” He stated that Herring became a gang member when he was 11 or 12 years old and that Herring’s substance abuse, drug trafficking, and other crimes were greatly influenced by his gang involvement.
{¶ 30} Herring also presented affidavits from his paternal grandmother, two aunts, two uncles, his sister Nicole, his mother, and a cousin. These affidavits discussed Herring’s gang involvement, his life as a drug dealer, his mother’s drug use, and other family members who have been incarcerated. Herring’s cousin and grandmother and an aunt and an uncle would have testified at trial if they had been asked.
{¶ 31} Herring’s sister, Nicole, stated that their grandmother took care of her and Herring until he was about 12, when their grandmother moved out of their home. Nicole indicated that their grandmother provided them with structure and discipline. She was strict and made Nicole and Herring do chores and complete their homework.
{¶ 32} Herring’s mother, Deborah, stated that she has seven children by four different fathers. Deborah said that she was overwhelmed with caring for her children. Herring’s father had been shot to death over a drug dispute when Herring was four years old. Deborah admitted that she has had a longstanding drinking problem and had been addicted to crack for 12 years. She knew that Herring sold crack but did not think that he had a substance-abuse problem.
2. Dr. Darnall’s letter
{¶ 33} Herring’s petition included a letter from Dr. Douglas Darnall, a clinical psychologist, that had been sent to trial counsel before the mitigation phase. At trial counsel’s request, Dr. Darnall had conducted pretrial testing of Herring by administering the Minnesota Multiphasic Personality Inventory (“MMPI-2”). Dr. Darnall told tidal counsel in the letter that he had not had the opportunity to *171conduct a clinical interview or a complete assessment and therefore was “unable to derive any specific clinical conclusions.”
{¶ 34} Dr. Darnall’s letter reported that the results of the MMPI-2 were of “questionable” validity. He stated that certain test results suggested that Herring was “exaggerating his symptoms or not understanding the specific test items.” Dr. Darnall added:
The configuration of the clinical scales would suggest that Mr. Herring has made some discriminations in his responses. The configuration would characterize Mr. Herring as a very suspicious individual who is likely to be hostile to authority figures. He is inclined to be rather impulsive, unreliable, egocentric, and irresponsible. Others may perceive him as suspicious, hostile, and hypersensitive to the reactions of others. * * * It is possible that Mr. Herring may have a delusional disorder that could be distorting his perceptions and interpretations as to what is going on around him. He may feel socially isolated and have persecutory ideas.
{¶ 35} Dr. Brams submitted an addendum affidavit that responded to Dr. Darnall’s report. Dr. Brams stated: “Even a cursory perusal of Mr. Herring’s MMPI-2 profile would indicate [the] need for further neuropsychological and psychiatric follow-up in terms of evaluation and testing.” She added: “Defense counsel did not follow-up this MMPI-2 report with any further evaluation of the defendant. Simply stated, the sole use of an MMPI * * * to determine mitigation factors or lack thereof is grossly limited and in no way could fully describe the functioning or history of that individual.”
3. Hrdy’s affidavit
{¶ 36} The affidavit of Thomas Hrdy, the mitigation specialist engaged by Herring’s attorneys prior to the trial, discussed his employment and his meeting with counsel in preparing for Herring’s mitigation:
I believe that I was first contacted by the trial attorneys Mr. Van Brocklin and Mr. Zena around the first week of August, 1997. As it turned out, this did not allow adequate time for me to collect the necessary records for a full investigation of the case, such as school records and medical records. I met with the attorneys only once, at our initial meeting. I do not know what information the attorneys collected. I do not know if the attorneys followed up on my suggestion to bring a psychologist in to evaluate Willie. I do not know these things because the *172attorneys did not share the information with me, and we did not have subsequent meetings after our initial one.
{¶ 37} In his affidavit, Hrdy stated that he had interviewed Herring four times and Herring’s mother once. He did not interview any other family members. Hrdy stated that due to time constraints, he was unable to complete extensive interview research. He did not recall “the specific attempts to collect specific records, such as Ohio Department of Youth Services or Mahoning County Human Services.” Hrdy added: “I felt like before I knew it, I was over my head in terms of time necessary to complete the investigation. I do not recall if I told trial counsel that I was running out of time.”
{¶ 38} Hrdy said that he had provided mitigation services “in two or three other capital cases” prior to Herring’s case. He concluded: “Looking back now on the work I did for the Herring case, I feel that I did a substandard job of mitigation investigation. This is primarily due to the fact that I underestimated the amount of time which was needed to contact all of the necessary mitigation witnesses and locate all the necessary resources and records.”
4. Hall’s affidavit
{¶ 39} In another affidavit attached to the petition, Dorian Hall, the supervisor of the mitigation section of the Office of the Ohio Public Defender, analyzed the deficiencies in Hrdy’s mitigation investigation. Hall stated that Hrdy obtained education records and attempted to obtain records from Southside Medical Center, “but did not appear to make any attempts to obtain additional records.” Hall also stated that Hrdy interviewed only Herring and his mother and neglected to interview various family members, friends, and professionals who could have provided detailed information about Herring’s life. Hall also stated that there was “a family history of substance abuse and drug dealing, gang involvement, criminal activity, neglect and lack of nurturing” that was not properly investigated.
B. Trial court denies Herring’s postconviction petition
{¶ 40} On January 6, 2003, the trial court granted the state’s motion for summary judgment and overruled Herring’s requests for discovery and an evidentiary hearing. The trial court rejected Herring’s claim that counsel were ineffective in conducting their mitigation investigation, stating:
It is clear from the transcript of the sentencing phase that counsel elected to present positive evidence from the Defendant’s family, and not to present negative evidence concerning the Defendant’s childhood. At *173this point, one can only speculate as to what effect, if any, negative evidence would have had in the jury’s deliberations. Tactical decisions and strategic choices must be reviewed with the strong presumption that effective legal counsel is rendered. * * * A different opinion, which varies from the theory used at trial does not depict ineffective assistance of counsel. * * *
In the instant case, Defendant simply suggests and speculates that trial counsels [sic] failure to present an alternative theory, specifically, negative testimony concerning his childhood, amounts to ineffective assistance of counsel. This Court does not agree, and the Defendant is not entitled to a hearing as to these claims.
{¶41} The trial court also rejected Herring’s claim that trial counsel were ineffective by failing to hire a neuropsychologist to examine him. The trial court noted that Dr. Darnall prepared and submitted a report and did “not recommend that the Defendant be examined by a neuro-psychologist. Additionally, there were specific findings made by Dr. Darnall that were not favorable to the Defendant.”
C. Court of appeals remands
{¶ 42} On appeal, the court of appeals held that the trial court had erred in granting summary judgment to the state. 7th Dist. Mahoning No. 03-MA-12, 2004-Ohio-5357, 2004 WL 2334325, ¶ 2, 115. The court reviewed the postconviction evidence that the defense presented in support of its ineffectiveness claim, id. at ¶ 70-91, and compared it with the meager amount of evidence that had been presented during mitigation, id. at ¶ 60-62. The court held that the trial court should not have rejected Herring’s arguments on the ground that trial counsel made a tactical decision to present only (minimal) positive mitigation evidence without first considering whether counsel had made an “intelligent strategic decision” after conducting a proper investigation. Id. at ¶ 99-100, citing Williams v. Taylor, 529 U.S. 362, 396, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000) (holding that counsel has the duty to conduct a thorough investigation of a capital defendant’s background). The appellate court further determined that “[t]here is little evidence herein documenting the extent of Appellant’s trial counsel’s] reasoning for their investigative decisions.” Id. at ¶ 114. The court observed that “Hrdy’s affidavit is inconclusive as to whether Appellant’s trial counsel actually knew of Hrdy’s investigation’s deficiencies.” Id. at ¶ 102.
{¶ 43} The court remanded the case to the trial court “to conduct an evidentiary hearing relative to Appellant’s trial counsel’s] efforts in advance of their decision to present only Appellant’s positive mitigation history.” Id. at ¶ 114. *174The court added: “Without a hearing to determine the extent of the mitigation evidence before Appellant’s trial counsel and their investigative efforts, Appellant’s postconviction exhibits may simply present an alternative mitigation tactic.” Id. at ¶ 104. In ordering an evidentiary hearing, the court specifically directed the trial court to “assess whether Appellant’s counsel were apprized of Hrdy’s investigation’s shortcomings. Only then could counsel have made a reasoned decision to cease investigating.” Id. at ¶ 116.
D. Trial court’s hearing on remand
{¶ 44} On August 28 and December 4, 2006, the trial court conducted an evidentiary hearing, at which Gary Van Brocklin and Thomas Zena, Herring’s trial attorneys, and Dorian Hall testified.. Documentary evidence was also presented. Hrdy did not testify at the hearing.
1. Van Brocklin’s testimony
{¶ 45} Van Brocklin had been lead counsel at Herring’s trial. Van Brocklin testified that after the defense had a difficult time obtaining a mitigation specialist, the Office of the Ohio Public Defender had identified Hrdy as a possible mitigation specialist. Hrdy was then contacted and hired as the defense mitigation specialist.
{¶ 46} Counsel’s first meeting with Hrdy in early September 1997 was four and one-half hours long. Van Brocklin recalled, “Mr. Hrdy was to obviously investigate and determine if we could find mitigation for Mr. Herring in the event that there was a conviction of capital specifications.” Van Brocklin could not recall whether defense counsel met with Hrdy again or only discussed the case over the phone.
{¶ 47} Van Brocklin stated that Hrdy never told defense counsel that “he in any way fell short of the mark” and always represented himself as an expert. Van Brocklin said, “I believed at the time that he had done all of the work necessary to look into Mr. Herring’s background. * * * I believe, that had Mr. Hrdy requested additional time, we would’ve immediately filed a motion for it, and I’m quite confident that Judge Durkin would’ve allowed us the time.”
{¶ 48} Van Brocklin also stated that Hrdy had had sufficient time to conduct an adequate mitigation investigation. Hrdy was hired in August or September 1997, and the start of the trial was delayed because a mistrial was declared during jury selection. The trial then eventually commenced on January 5, 1998, and the mitigation proceedings began on February 14,1998.
{¶ 49} Van Brocklin testified that he and Zena “knew a lot of negative information” about Herring “through our own investigation and through criminal records and those kinds of things that were supplied to us during the lengthy *175discovery process.” Van Brocklin could not recall specific information that they had known about Herring’s background because the case files had been transferred to the public defender’s office. Yet Van Brocklin had known that Herring had a prior juvenile record and that members of his family had criminal records. Van Brocklin also recalled that a subpoena had been issued for Herring’s confinement records to show that he had behaved himself during incarceration. But defense counsel decided not to use that information because the records showed Herring’s involvement in “a death threat or a fight.”
{¶ 50} Van Brocklin also stated that the defense had hired Dr. Darnall. Dr. Darnall provided defense counsel with “a report that didn’t work well, or was not very definitive.” He did not perform any testing for neurological impairment.
{¶ 51} Van Brocklin stated that he and Zena decided to present “positive mitigation evidence.” He also stated, “I hammered home in argument * * * that Mr. Herring had not been convicted as a principal offender in this matter.” Van Brocklin explained that the decision to present positive evidence in mitigation had “a lot to do with the fact [than] we had picked two juries, and the second jury was far more conservative that the first jury panel was, and both Tom and I did not think that negative information would have worked with that particular jury panel.” He added, “I thought that any kind of information that you would give the second jury panel that Mr. Herring had been involved in a life of crime would simply be more ammunition for them to find a death verdict.”
2. Zena’s testimony
{¶ 52} Zena had been co-counsel at Herring’s trial and had the primary responsibility for preparing mitigation. Zena testified that the defense initially was prepared to engage a different mitigation specialist, but after that person could not take the job, trial counsel hired Hrdy. Zena did not believe that they developed a theory of mitigation with Hrdy. Zena stated, “I think we said let’s see what we have and try to put the best foot forward to save [Herring], whatever that might be.” As to Hrdy’s role, Zena said that although “it’s the lawyer’s responsibility to present the mitigation,” people in Hrdy’s position are called mitigation specialists because it is their job to “go out and find and give ideas on how they think something should be presented.”
{¶ 53} Zena stated that he and Van Brocklin had talked to Hrdy on the phone as Hrdy’s mitigation investigation progressed. Counsel did not have a timetable for getting information from Hrdy. Zena testified:
[I]t was like if you need anything, call us, and we had our normal discussions. If there’s anything you’re having trouble with or need, call us. And I don’t think there was ever a time when we weren’t accessible to *176him. I know every time he tried to reach me, I was. Likewise, he was accessible.
{¶ 54} Zena testified that he had met Herring’s mother on a few occasions and got to know Herring’s sister. Zena stated that he spoke to Herring’s mother about “arranging a meeting with anybody she thought would be helpful with mitigation, and we met at their home.” Herring’s mother never told Zena that she had been a drug addict during a large part of Herring’s life.
{¶ 55} Zena also had spoken to Herring about mitigation. He stated that Herring told him nothing negative about the family, because Herring “is a person that did not accent the negative about any family member or himself.” Zena stated, “When we got to the mitigation phase, he was cooperative in the sense of knowing we — what we were going to do, but he was not forthcoming with any information.” But Herring did not put any restrictions on trial counsel’s presentation of mitigation evidence.
{¶ 56} Zena stated that he and Van Brocklin made a “conscious choice to * * * put on good things about him” and to emphasize that Herring was not a principal offender. Zena explained that this was “an awful case as cases go. * * * This was mayhem in a bar where people wound up dead, people wound up shot, bullets all over the floor.” Moreover, the victims were innocent bystanders who were not “involved in any transgressions with any of the individuals who came in.” Thus, Zena said that the goal was “separating [Herring] out as a non principal offender, not responsible for the death of anyone, and showing whatever I could of the other side of him.”
3. Billing statements
{¶ 57} Defense exhibit A was Van Brocklin’s billing statement, submitted in February 1998, for serving as Herring’s defense counsel at trial. Hrdy’s billing statement and its cover letter were attached. Hrdy’s letter included the observation, “This has been a most difficult case to find mitigation on as you well know[.] I know you did the best you could with the little I provided.” Hrdy’s billing statement showed that he sent a bill to counsel for time and expenses totaling $1,501.30.
4. Hall’s testimony
{¶ 58} Hall explained the responsibilities of a mitigation specialist in conducting an investigation. Hall stated that Hrdy’s invoice showed that he had completed approximately 30 hours of work. Hall did not believe that this was an adequate amount of time to properly conduct a mitigation investigation for Herring.
*177 E. Trial court again denies Herring’s petition
{¶ 59} On September 26, 2008, the trial court again denied Herring’s petition for postconviction relief.
{¶ 60} After reviewing the evidence presented at the hearing, the trial court made the following ruling:
Based on this evidence, it is abundantly clear that Thomas Hrdy never advised trial counsel that his investigation was not complete, and never asked them for additional time to complete it. Trial counsels [sic] decision to present positive mitigation was reasonable, based on an objective review of counsels [sic] performance, measured with reasonableness under professional norms, including a context-dependent consideration of the challenged conduct as seen from counsels [sic] perspective at the time of that conduct.
In addition, consistent with the Trial Court’s opinion that granted summary judgment to the State, “one can only speculate as to what effect, if any, negative evidence would have had in the jury’s deliberations.”
F. Court of appeals orders a new sentencing hearing
{¶ 61} On appeal following this ruling, the court of appeals held that the trial court abused its discretion in finding that trial counsel’s decision to present only positive mitigation evidence was reasonable. 7th Dist. Mahoning No. 08-MA-213, 2011-Ohio-662, 2011 WL 497765, ¶ 90. The court stated: “Trial counsel can make the decision to forego the presentation of evidence, but only after a full investigation. * * * Only after completing a full investigation can counsel make an informed, tactical decision about what information to present in their client’s case.” Id. at ¶ 55, citing Wiggins v. Smith, 539 U.S. 510, 521, 123 S.Ct. 2527, 156 L.Ed.2d 471 (2003), and Williams, 529 U.S. 362, 120 S.Ct. 1495, 146 L.Ed.2d 389.
{¶ 62} The court stated that information about Herring’s background in the postconviction affidavits
brought to light appellant’s deeply troubled childhood, his complete lack of any positive role models, his substance abuse problems, his depression, his low IQ, and his possible organic brain impairment. These areas of appellant’s life, had they been investigated and explored fully, are all very significant factors to be weighed and considered in determining what mitigation evidence to present. And counsel did not have this information *178before them when they made the decision to present only positive mitigation evidence.
Id. at ¶ 79. The court then concluded that “counsel could not have made an intelligent strategic decision without the proper investigation before them.” Id.
{¶ 63} The court also stated that investigations into mitigating evidence “ ‘ “should comprise efforts to discover all reasonably available mitigating evidence.” ’ ” (Emphasis sic.) Id. at ¶ 82, quoting Wiggins at 524, quoting American Bar Association (“ABA”) Guidelines for the Appointment and Performance of Counsel in Death Penalty Cases, Guideline 11.4.1(C) (1989). The court determined that trial counsel’s testimony showed that they had failed to meet this standard. Id. at ¶ 82. The court also noted that Hrdy admitted that his investigation was “substandard” and that he did not complete many of the tasks that he should have in investigating Herring’s background. Id. at ¶ 79. But the court stated that it was trial counsel’s duty to ensure that a complete investigation was conducted and that they could not simply rely on Hrdy’s investigation. Id. at ¶ 82.
{¶ 64} In ordering a new sentencing hearing, the court of appeals stated that the undiscovered mitigating evidence “ ‘ “might well have influenced the jury’s appraisal” ’ ” of Herring’s culpability. Id. at ¶ 90, quoting Rompilla v. Beard, 545 U.S. 374, 393, 125 S.Ct. 2456, 162 L.Ed.2d 360 (2005), quoting Williams, 529 U.S. at 398. The court of appeals also stated that “the probability of a different sentence if counsel had presented the evidence is ‘ “sufficient to undermine confidence in the outcome” ’ reached by the jury.” Id., quoting Rompilla at 393, quoting Strickland, 466 U.S. at 694, 104 S.Ct. 2052, 80 L.Ed.2d 674.
IV. Issues on appeal
{¶ 65} The cause is now before this court following the acceptance of the state’s discretionary appeal on the following three propositions of law:
I. Defense counsel’s performance is constitutionally effective under the federal and state Constitutions where, absent any knowledge of a mitigation expert’s shortcomings, they proceed reasonably in light of the information that they have obtained, and despite the fact that a mitigation expert failed to complete several tasks in preparation for the sentencing phase (of a capital trial).
II. Capital defendants do not have a federal constitutional right to the effective assistance of a mitigation specialist; therefore, a mitigation *179specialist’s deficiencies cannot be imputed to trial counsel without having sufficient knowledge of those deficiencies.
III. An appellate court errs in finding that trial counsel was constitutionally ineffective without determining whether or not the defendant suffered actual prejudice as a result of trial counsel’s performance, as set forth in Strickland v. Washington.
Y. Analysis
A. Deficient mitigation investigation (Proposition of Law I)
{¶ 66} The state argues that trial counsel were not deficient in presenting positive mitigating information about the defendant. The state contends that trial counsel performed reasonably even though the defense mitigation specialist failed to conduct a complete investigation, because counsel were unaware of those shortcomings.
1. Legal standards
{¶ 67} To establish a violation of the Sixth Amendment right to counsel, Herring must demonstrate that counsel’s performance was deficient and that the deficient performance prejudiced him. Strickland, 466 U.S. at 687, 104 S.Ct. 2052, 80 L.Ed.2d 674.
{¶ 68} In assessing counsel’s investigation, an objective review of counsel’s performance must be conducted in light of professional norms prevailing when the representation took place. Bobby v. Van Hook, 558 U.S. 4, 7, 130 S.Ct. 13, 175 L.Ed.2d 255 (2009); Strickland at 688. Under the deficient-performance prong, the court should “indulge a strong presumption that counsel’s conduct falls within the wide range of reasonable professional assistance.” Id. at 689.
{¶ 69} Counsel in a capital case have an “obligation to conduct a thorough investigation of the defendant’s background” to determine the availability of mitigating evidence. Williams, 529 U.S. at 396, 120 S.Ct. 1495, 146 L.Ed.2d 389. Counsel’s “investigations into mitigating evidence ‘should comprise efforts to discover all reasonably available mitigating evidence and evidence to rebut any aggravating evidence that may be introduced by the prosecutor.’ ” (Emphasis sic.) Wiggins, 539 U.S. at 524, 123 S.Ct. 2527, 156 L.Ed.2d 471, quoting ABA Guidelines for the Appointment and Performance of Counsel in Death Penalty Cases, Guideline 11.4.1(C).
*180This constitutionally required background investigation is necessary to enable counsel to make strategic choices about presenting a mitigation defense. * * * Indeed, the deference owed to counsel’s strategic judgments about mitigation is directly proportional to the adequacy of the investigations supporting such judgments. Accordingly, when evaluating the reasonableness of counsel’s mitigation strategy in a capital case, “a reviewing court must consider the reasonableness of the investigation said to support that strategy.”
Jells v. Mitchell, 538 F.3d 478, 492 (6th Cir.2008), quoting Wiggins at 527.
2. Discussion
{¶ 70} The state argues that defense counsel proceeded reasonably in presenting positive mitigating evidence about Herring to the jury. The state recognizes that Wiggins required that a mitigation investigation be completed. The state argues, however, that it was professionally reasonable for counsel to believe that Hrdy had completed his investigation prior to sentencing and that defense counsel did not have to ensure for themselves that the mitigation investigation was accurate and complete.
(a) Inadequate investigation
{¶ 71} The affidavits and testimony presented during postconviction proceedings show that neither Hrdy nor trial counsel obtained detailed information about Herring’s background. Thus, trial counsel did not have detailed knowledge about Herring’s parental neglect, gang involvement, or life as a drug dealer. Trial counsel also does not appear to have known that Herring’s father had been murdered when Herring was a young child or that his mother had been a drug addict during a large part of her life.
{¶ 72} Hrdy, the mitigation specialist, admits that he did a “substandard job of mitigation investigation.” He acknowledges that he failed to interview family members. Hrdy also failed to obtain Herring’s DYS records, police records, or other records that might have shed light on Herring’s background. It appears that Hrdy obtained only Herring’s education records and attempted to obtain part of his medical records. Hrdy’s billing statement also suggests that he did not put in the time necessary to conduct an adequate investigation.
{¶ 73} Moreover, Herring had not been evaluated and tested to explore the psychological, developmental, and intellectual factors in his background. Dr. Brams conducted IQ and other psychometric tests that uncovered a wealth of information about Herring.
*181{¶ 74} The state responds to these concerns by pointing out that defense counsel employed Dr. Darnall, a clinical psychologist, and obtained a report from him. But Dr. Darnall administered only the MMPI-2 to Herring. Dr. Darnall reached some tentative conclusions about Herring, including observing: “It is possible that Mr. Herring may have a delusional disorder that could be distorting his perceptions and interpretations as to what is going on around him.” Yet defense counsel did not arrange for follow-up psychological evaluations or other testing of Herring.
{¶ 75} Herring’s trial counsel, Van Brocklin, testified that he and Zena “knew a number of negative things through our own investigation and through criminal records and those kinds of things that were supplied to us during the lengthy discovery process.” Yet Van Brocklin and Zena were unable to recall many specific details about their knowledge because they no longer had access to Herring’s case file.
{¶ 76} Here, the mitigation investigation was less comprehensive than the investigation that the Supreme Court found deficient in Wiggins. In that case, defense counsel had arranged for a psychologist to run a number of tests on the defendant. 539 U.S. at 523, 123 S.Ct. 2527, 156 L.Ed.2d 471. Counsel also had “had available to them a written PSI [presentence-investigation report], which included a one-page account of Wiggins’ ‘personal history’ noting his ‘misery as a youth,’ quoting his description of his own background as ‘ “disgusting,” ’ and observing that he spent most of his life in foster care.” Id. Counsel also had tracked down records kept by the Baltimore City Department of Social Services (“DSS”) that documented various placements in foster care. Id. The PSI and the DSS records revealed that Wiggins’s mother was a chronic alcoholic, Wiggins had been “shuttled from foster home to foster home and displayed some emotional difficulties,” he had frequent, lengthy absences from school, and “on at least one occasion, his mother left him and his siblings alone for days without food.” Id. at 525. Despite these leads, counsel did not investigate further. The Supreme Court remarked that “any reasonably competent attorney would have realized that pursuing these leads was necessary to making an informed choice among possible defenses.” Id. Had counsel investigated further, the court stated, they might well have discovered the severe physical and sexual abuse that the defendant had suffered from his mother and while in the care of a series of foster parents. Id. at 516.
{¶ 77} In Porter v. McCollum, 558 U.S. 30, 40, 130 S.Ct. 447, 175 L.Ed.2d 398 (2009), the Supreme Court held that counsel’s failure to interview the defendant’s family members and to obtain school, medical, and military service records made his representation constitutionally deficient. The court stated that “like the counsel in Wiggins, [defense counsel] ignored pertinent avenues for investigation *182of which he should have been aware,” and it concluded that the “decision not to investigate did not reflect reasonable professional judgment.” Id.
{¶ 78} Moreover, this is not a case like Van Hook, in which the Supreme Court held that counsel were not deficient in failing to dig deeper into the defendant’s background. In Van Hook, the defense had called eight mitigation witnesses and the defendant had made an unsworn statement. 558 U.S. at 5, 130 S.Ct. 13, 175 L.Ed.2d 255. As for the mitigation investigation, the court noted that defense counsel had contacted their lay witnesses “early and often” before trial. Id. at 9. Counsel spoke nine times with the defendant’s mother, twice with an aunt who lived with the family and often cared for Van Hook as a child, and three times with a family friend. Id. They also were in touch with their two expert witnesses and reviewed the defendant’s military records. After reviewing his military records, they met with a representative from the Veterans’ Administration seven weeks before trial. Id.
{¶ 79} Despite all the evidence that defense counsel had uncovered and presented, Van Hook argued that his counsel were deficient by not interviewing other family members — his stepsister, two uncles, and two aunts — as well as a psychiatrist who had once treated his mother. Id. at 11. In rejecting this claim, the Supreme Court stated that “there comes a point at which evidence from more distant relatives can reasonably be expected to be only cumulative.” Id. Thus, the court concluded, “it was not unreasonable for his counsel not to identify every other living family member or every therapist who once treated his parents.” Id.
{¶ 80} The investigation in this case was incomplete because counsel failed to interview witnesses and obtain records about Herring’s dysfunctional childhood, gang involvement, and substance abuse, and about his mother’s drug addiction. Defense counsel also failed to ensure that Herring was adequately evaluated and tested by a psychologist. Thus, this is not a case in which information about Herring’s background that was not uncovered would have been merely cumulative.
{¶ 81} Moreover, trial counsel’s decision not to expand their investigation fell short of the standards for capital defense work articulated by ABA guidelines. The Supreme Court of the United States has stated that the ABA standards that were in effect at the time of trial are guides to determining what is reasonable. Van Hook, 558 U.S. at 8-9, 130 S.Ct. 13, 175 L.Ed.2d 255. The 1989 ABA guidelines that were in effect at the time of Herring’s trial called for Herring’s counsel “to discover all reasonably available mitigating evidence and evidence to rebut any aggravating evidence that may be introduced by the prosecutor.” ABA Guidelines for the Appointment and Performance of Counsel in Death Penalty Cases, Guideline 11.4.1(C).
*183{¶ 82} The 1989 guidelines provided detailed guidance as to counsel’s investigative responsibilities. They stated that “[a]s soon as is appropriate,” counsel should
[cjollect information relevant to the sentencing phase of trial including, but not limited to: medical history (mental and physical illness or injury of alcohol and drug use, birth trauma and developmental delays); educational history (achievement, performance and behavior)[;] special educational needs (including cognitive limitations and learning disabilities); military history (type and length of service, conduct, special training); employment and training history (including skills and performance, and barriers to employability); family and social history (including physical, sexual or emotional abuse); prior adult and Juvenile record; prior correctional experience (including conduct or supervision and in the institution/education or training/elinical services); and religious and cultural influences.
Id. at Guideline 11.4.1(D)(2)(C).
{¶ 83} Counsel’s performance fell below these well-defined norms. Counsel failed to ensure that a comprehensive investigation was conducted into Herring’s background, obtaining only rudimentary information about Herring’s childhood, substance abuse, gang involvement, and psychological makeup. Based on these failures, we conclude that counsel did not conduct an adequate investigation into Herring’s background before the mitigation hearing started.
(b) Defense strategy to introduce positive mitigation
{¶ 84} The state argues that trial counsel made a “strategic decision” to present positive mitigating evidence. Van Brocklin testified that they decided to present only positive mitigation because some of the jurors appeared to lean in favor of the death penalty. He stated that “any kind of information * * * that Mr. Herring had been involved in a life of crime would simply be more ammunition for them to find a death verdict.”
{¶ 85} The state also points to Van Brocklin’s statement that trial counsel “knew a number of negative things” about Herring and wanted to keep them from the jury. Although his recollection was imperfect, Van Brocklin testified that they knew about Herring’s juvenile record and that members of Herring’s family had criminal records. Trial counsel also employed Dr. Darnall, who conducted some pretrial testing of Herring, though the scope of Dr. Darnall’s evaluation was quite limited.
{¶ 86} The state invokes Burger v. Kemp, 483 U.S. 776, 107 S.Ct. 3114, 97 L.Ed.2d 638 (1987), for the proposition that counsel’s decision to keep negative *184information from the jury was a sound trial strategy and did not necessitate a full mitigation investigation. In Burger, the Supreme Court upheld the propriety of a defense decision not to present any mitigating evidence on behalf of the defendant. The court rejected claims that defense counsel was deficient by not pursuing an all-out investigation into the defendant’s background in search of mitigating evidence. Id. at 794. The court stated that the limited investigation conducted was reasonable because counsel “did interview all potential witnesses who had been called to his attention,” and discovered little that was helpful and much that was harmful. Id. .at 794-795.
{¶ 87} The state also cites Scott v. Mitchell, 209 F.3d 854 (6th Cir.2000), in arguing that trial counsel could reasonably have presented only positive mitigation without conducting a full mitigation investigation. In Scott, defense counsel pursued a residual-doubt strategy and presented no mitigating evidence other than Scott’s unsworn statement to the jury. Id. at 880. In state postconviction proceedings, the defense presented evidence showing that counsel had failed to contact family members who would have told them that the defendant’s parents were alcohol and drug abusers, the defendant grew up in severe poverty, and he was exposed to an exceedingly violent environment during his upbringing. See State v. Scott, 63 Ohio App.3d 304, 309-311, 578 N.E.2d 841 (8th Dist.1989).
{¶ 88} The Sixth Circuit Court of Appeals stated: “Scott’s penalty-phase attorneys would certainly have been well-advised to conduct more research into mitigating factors than they did.” 209 F.3d at 881. But the court declined to find that counsel were ineffective by failing to conduct a more thorough investigation. The court noted in dicta that counsel’s decision to pursue a residual-doubt strategy in the case was not unreasonable when counsel had conducted research into the available mitigating testimony and wanted to avoid opening the door to evidence of the defendant’s extensive criminal history. Id. at 880-882.
{¶ 89} Unlike counsel in Burger, Herring’s counsel did not talk to every witness who was brought to their attention and did not have a psychologist conduct a comprehensive psychological evaluation of the defendant. Trial counsel also did not review Herring’s DYS records and other records that would have provided information about his dysfunctional background. As for the applicability of Scott, the Sixth Circuit has since held that incomplete mitigation investigations like those in that case are legally inadequate. In Goodwin v. Johnson, 632 F.3d 301 (6th Cir.2011), defense counsel had presented a residual-doubt strategy like that in Scott. Counsel had failed, however, to speak to most of Goodwin’s relatives, had not reviewed his school records, and had not had him evaluated. Id. at 324. The court stated that had counsel interviewed these witnesses and examined school and juvenile records, they would have learned that he was abused by his drug-using mother, performed poorly in school, and had psyeholog*185ical problems. Id. The court held that counsel’s decision to forgo presenting mitigating evidence was not an informed decision and that counsel’s performance was inadequate. Id. at 325-326. See also Foust v. Houk, 655 F.3d 524, 534-536 (6th Cir.2011) (partial but incomplete mitigation investigation deemed inadequate); Mason v. Mitchell, 543 F.3d 766, 780 (6th Cir.2008) (same); Jells, 538 F.3d at 496 (same).
{¶ 90} Herring acknowledges that a decision by trial counsel to present only positive mitigation can be a sound trial strategy “in the right case.” Herring cites Wiggins in arguing that trial counsel’s decision to pursue a positive-mitigation theory can properly be made only after counsel has conducted a full mitigation investigation. We agree. Thus, counsel’s decision to pursue a positive-mitigation theory was not justified because it was made before an adequate investigation had been conducted into Herring’s background.
(c) Trial counsel’s responsibility for the investigation
{¶ 91} The state argues that trial counsel were not ineffective, because they were unaware that Hrdy had failed to conduct an adequate investigation. Thus, the state asserts that it was professionally reasonable for counsel to believe that Hrdy had completed his investigation and to rely on his status as a mitigation expert rather than ensuring for themselves that the investigation was accurate and complete.
{¶ 92} The state cites Drummond v. Houk, 761 F.Supp.2d 638 (N.D.Ohio 2010), aff'd on other grounds, 728 F.3d 520 (6th Cir.2013), vacated and remanded sub nom. Robinson v. Drummond, — U.S.-, 134 S.Ct. 1934, 188 L.Ed.2d 957 (2014), in arguing that trial counsel were not responsible for Hrdy’s failure to conduct an adequate investigation. In Drummond, Dr. John Fabian, a clinical psychologist, at the penalty phase of the trial provided mitigating testimony about the defendant’s background and gang involvement. Id. at 695-696. But later during mitigation, Dr. Fabian testified that he was “not a gang expert,” when he was questioned about the environmental factors in a gang atmosphere. Id. at 696. Dr. Fabian also provided other testimony about Drummond’s gang involvement that was not helpful to the defense. Id. at 696-697.
{¶ 93} During federal habeas proceedings, Drummond claimed that trial counsel had been ineffective for failing to hire a “gang expert,” for failing to prepare Dr. Fabian for testimony, and for failing to conduct a complete investigation. Id. at 702. During the proceedings, Dr. Fabian stated that he had not had time to prepare for trial and had been unprepared to testify. Id. at 697, 699-700. Defense counsel responded that Dr. Fabian had not expressed concerns to them about having had insufficient time to prepare his testimony. Id. at 703. Defense counsel also stated that based upon Dr. Fabian’s representations regarding his *186credentials and experience, they had believed that Dr. Fabian could testify as effectively as any “gang expert.” Id. at 702.
{¶ 94} In rejecting Drummond’s ineffectiveness claims, the court stated that “it was reasonable for counsel to rely on Dr. Fabian’s self-described experience with gang members and to presume that Dr. Fabian had sufficient time to prepare for his trial testimony absent any credible assertions to the contrary.” Id. at 703. The court concluded that “counsel cannot be held responsible for the misrepresentations or omissions of Dr. Fabian.” Id.
{¶ 95} The state argues that as in Drummond, Hrdy never told defense counsel that his investigation was incomplete or that he needed more to time to complete it. Thus, the state argues that defense counsel cannot be held responsible for Hrdy’s misrepresentations or omissions. There are important differences between the two cases. In Drummond, the court focused on the testimony of a single witness. In contrast here, Hrdy, the mitigation specialist, was responsible for a full investigation of the case by collecting records, interviewing witnesses, and providing an expert-witness list.
{¶ 96} The state also argues that trial counsel could reasonably have relied on Hrdy’s expertise as a mitigation expert. Van Brocklin testified that the defense obtained Hrdy’s name through the state public defender’s office and that Hrdy “always represented himself as an expert.” Trial counsel’s reliance on Hrdy’s expertise as a mitigation specialist was not the same as counsel’s reliance on Dr. Fabian’s expertise in Drummond. -Dr. Fabian was a clinical psychologist and counsel should be able to rely on a psychologist’s professional representations. See Clark v. Mitchell, 425 F.3d 270, 285 (6th Cir.2005) (it was not unreasonable for counsel, untrained in the field of mental health, to rely on the opinion of a clinical psychologist).
{¶ 97} Additionally regarding the state’s claim that counsel could reasonably have relied on Hrdy’s expertise, it is instructive that the 1989 ABA guidelines did not mention mitigation specialists. Not until the 2003 edition do the ABA guidelines call for the hiring of a mitigation specialist as part of the defense team. See ABA Guidelines for the Appointment and Performance of Defense Counsel in Death Penalty Cases, Guidelines 4.1(A)(1) and 10.4(C)(2)(a) (Rev.Ed.2003). Even then, the guidelines call for the mitigation specialist to serve in an investigatory and advisory capacity. See, e.g., id. at Guideline 10.4(B) (“counsel bears overall responsibility for the performance of the defense team”). Thus, we reject the state’s argument that counsel could rely on Hrdy’s representations that he was a mitigation expert without conducting further investigation.
{¶ 98} The state also argues that it was reasonable for defense counsel to believe that Hrdy had completed his investigation because of the amount that Hrdy billed for his work. Hrdy stated in his initial letter to counsel that he *187rarely needed more than $2,500 to complete an investigation, and he sent a bill after trial to counsel for a total of $1,501.30. It is unclear when trial counsel received Hrdy’s bill. But the billing represented only a little more than 60 percent of Hrdy’s expected expenditures. A review of the billing statement also shows that Hrdy had one meeting with Herring’s mother and four meetings with Herring and did not meet with any other family members or potential mitigation witnesses. Thus, contrary to the state’s argument, Hrdy’s billing statement put counsel on notice that Hrdy’s investigation had been incomplete and more work should have been done.
{¶ 99} In Johnson v. Bagley, 544 F.3d 592 (6th Cir.2008), the Sixth Circuit addressed similar ineffectiveness claims in a capital case. Counsel’s mitigation strategy in that case had been to humanize the defendant through his grandmother’s testimony and to present her as a compelling witness who would suffer from a jury decision to impose a death sentence. A central theme of the defense strategy was to present the grandmother as a pivotal figure in the defendant’s life, who “did everything that one could reasonably expect to do to try [to] help” him. Id. at 599-600.
{¶ 100} The court stated that “[i]n the abstract,” counsel’s mitigation strategy might have been a “legitimate strategic decision.” Id. at 600. But the court determined that trial counsel “pursued this strategy after what can only be described as an anemic and leaderless investigation” that suffered from significant flaws. Id.
{¶ 101} First, the court stated that the defense team chose not to interview the defendant’s mother because she had been a prostitute and a drug addict and would have been a “bad mitigation witness.” Id. The court stated that the mother’s bad background was precisely the reason why she should have been interviewed. Id. Second, the court stated that the defense obtained a large number of files from the Ohio Department of Human Services but apparently never read them. Instead, defense counsel simply submitted them to the jury without knowing whether they hurt the defendant’s strategy or helped it. Id. The court stated that if counsel had read the records they would have learned that social workers were reluctant to place the defendant in his grandmother’s custody because of her-abusive history. Thus, a review of the records would have tipped off counsel to a different mitigation theory and avoided the pitfall of submitting records to the jury that directly contradicted counsel’s theory that the defendant’s grandmother was a positive force for change in his life. Id. at 600-601.
{¶ 102} In a situation with similarities to the present case, Johnson’s trial counsel acknowledged that they were not involved in the mitigation investigation. Trial counsel had provided their mitigation specialists with an initial set of names, *188but one attorney admitted that he had provided no significant guidance, saying, “I don’t plan the investigation. We get the mitigation experts out to do that.” 544 F.3d at 601. On the day before trial began, that attorney admitted that he did not know what his investigator had done, and midway through the trial, defense counsel did not “even know what records [they were] going to have.” Id.
{¶ 103} The Sixth Circuit stated that these investigative blunders occurred because no one who participated in Johnson’s penalty-phase defense made any deliberate decisions about the scope of the investigation. Id. The Sixth Circuit noted that the lack of structure and supervision over the investigation led to significant delays and missed appointments and “ ‘suggests] that [the defense’s] incomplete investigation was the result of inattention, not reasoned strategic judgment.’ ” Id. at 602, quoting Wiggins, 539 U.S. at 534, 123 S.Ct. 2527, 156 L.Ed.2d 471. The court concluded that defense counsel’s presentation could not be justified as a strategic decision since they “were not in a position to make * * * reasonable strategic choice[s] * * * because the investigation supporting their choice[s] was unreasonable.” Id. at 603, quoting Wiggins at 536, 123 S.Ct. 2527.
{¶ 104} As in Johnson, Herring’s trial counsel had the responsibility to ensure that a complete investigation into Herring’s background was performed before the penalty phase. Johnson also demonstrates that trial counsel cannot avoid their responsibility for the failure to complete an adequate investigation into Herring’s background because of Hrdy’s failures. Moreover, as in Johnson, the evidence shows that trial counsel’s failure to complete the investigation resulted from “inattention, not reasoned strategic judgment.” Id. at 602. We conclude that Hrdy’s shortcomings did not excuse counsel’s failure to ensure that an adequate investigation was completed.
(d) Herring’s lack of cooperation
{¶ 105} The state also argues that Herring’s refusal to discuss or divulge any negative information about himself or his family precludes him from claiming that counsel’s investigation was inadequate.
{¶ 106} Zena testified that Herring had not been forthcoming with any negative information about his family. But Zena stated that Herring did not place any restrictions on the “presentment of the mitigation.” Zena explained, “When we got to the mitigation phase, [Herring] was cooperative in the sense of knowing * * * what we were going to do, but he was not forthcoming with any information.” Herring’s lack of cooperation in preparing for mitigation is an important factor in reviewing whether counsel was deficient.
{¶ 107} To determine whether counsel’s performance was deficient, the court must measure it against an objective standard based on accepted professional *189norms. See Rompilla, 545 U.S. at 380, 125 S.Ct. 2456, 162 L.Ed.2d 360. As a starting point, neither Wiggins nor Strickland addresses a situation in which a defendant fails to cooperate with counsel’s efforts to present mitigating evidence to a sentencing court. In Rompilla, the defendant refused to assist counsel in the development of a mitigation case, id. at 381, but there is no indication that the defendant ever informed the court or his counsel that he did not want mitigating evidence presented. The Supreme Court held that counsel was responsible for conducting a further investigation even though the defendant had suggested to counsel that no mitigation was available. Id. at 381-389.
{¶ 108} In some extreme cases, courts have held that counsel’s investigation was reasonable given the defendant’s lack of cooperation. In Schriro v. Landrigan, 550 U.S. 465, 127 S.Ct. 1933, 167 L.Ed.2d 836 (2007), the defendant actively obstructed counsel’s investigation and outright refused to allow counsel to present mitigating evidence. For example, the defendant explicitly instructed his mother and ex-wife not to testify and they refused to do so. Id. at 469-470. Counsel tried to make a proffer of the witnesses’ testimony, but the defendant repeatedly interrupted his counsel’s presentation to the court to reiterate that he did not want mitigating evidence presented. Id. at 470. The Supreme Court held that the defendant’s refusal to cooperate in the penalty phase rendered counsel’s limited investigation and presentation of mitigating evidence reasonable under the circumstances. Id. at 475-477.
{¶ 109} In Owens v. Guida, 549 F.3d 399 (6th Cir.2008), the court considered whether counsel’s failure to investigate a capital defendant’s background was deficient when the defendant would not cooperate. In Owens, the defendant would not cooperate with mental-health examiners, would not allow counsel to communicate with her family and, contrary to counsel’s advice, would not take the stand herself. Id. at 406-407. The court held that any failure to develop mitigating evidence was the result of the defendant’s actions and not deficient performance by her counsel. Id. at 412. The court stated, “A defendant cannot be permitted to manufacture a winning [ineffective-assistance-of-counsel] claim by sabotaging her own defense, or else every defendant clever enough to thwart her own attorneys would be able to overturn her sentence on appeal.” Id.
{¶ 110} Herring’s refusal to cooperate appears to fall somewhere between Rompilla and Landrigan. Herring was not forthcoming with counsel during mitigation, particularly with any negative information about his family. But Herring never told trial counsel that he did not want mitigation evidence presented. Thus, Herring’s refusal to divulge information did not excuse counsel *190from conducting a mitigation investigation. See State v. Neyland, 139 Ohio St.3d 353, 2014-Ohio-1914, 12 N.E.3d 1112, ¶ 242-249.
3. Conclusion: Counsel’s mitigation preparation was deficient
{¶ 111} Trial counsel’s strategy to present only “positive mitigation” was deficient because neither defense counsel nor the mitigation specialist completed a thorough mitigation investigation beforehand. Trial counsel’s responsibility to ensure that an investigation was completed cannot be excused because of Hrdy’s omissions, Hrdy’s failure to communicate with counsel, or Herring’s refusal to provide trial counsel with negative information about his family.
B. No constitutional right to mitigation specialist; imputation of Hrdy’s deficiencies to counsel (Proposition of Law II)
{¶ 112} The state’s second proposition recasts its claim that trial counsel were not responsible for inadequacies in Hrdy’s investigation. The state argues that Hrdy’s deficiencies in completing the investigation cannot be imputed to counsel, who did not have sufficient knowledge of them, because capital defendants do not have a constitutional right to the effective assistance of a mitigation specialist.
{¶ 113} It is true that Herring did not have a constitutional right to a mitigation specialist or a right to an effective one. See Moore v. Mitchell, 708 F.3d 760, 777 (6th Cir.2013); State v. McGuire, 80 Ohio St.3d 390, 399, 686 N.E.2d 1112 (1997) (no requirement for counsel to hire mitigation specialist in capital case). Even though “counsel did not have a specific obligation to employ a mitigation specialist, they did have an obligation to fully investigate the possible mitigation evidence available.” Jells, 538 F.3d at 495. The state’s argument that trial counsel’s responsibilities were lessened because Herring did not have a constitutional right to an effective mitigation specialist is unpersuasive. As discussed regarding the state’s first proposition, trial counsel had the responsibility to ensure that a complete investigation was performed before making the decision to present only positive mitigating evidence. Moreover, trial counsel cannot avoid their responsibility by relying on Hrdy’s failure to tell them that his investigation was incomplete. That is particularly true in this case, in which the facts show that the inadequate investigation resulted from trial counsel’s inattention and failure to monitor Hrdy’s progress in conducting it.
C. Prejudice inquiry (Proposition of Law III)
{¶ 114} The court of appeals concluded that trial counsel’s deficient performance was prejudicial. The court stated that the undiscovered mitigating evidence “ ‘ “might well have influenced the jury’s appraisal” ’ ” of Herring’s culpability. 2011-Ohio-662, 2011 WL 497765, at ¶ 90, quoting Rompilla, 545 U.S. at 393, 125 S.Ct. 2456, 162 L.Ed.2d 360, quoting Williams, 529 U.S. at 398, 120 S.Ct. 1495, *191146 L.Ed.2d 389. The court also stated that “the probability of a different sentence if counsel had presented the evidence is ‘ “sufficient to undermine the confidence in the outcome.” ’ ” Id., quoting Rompilla at 393, quoting Strickland, 466 U.S. at 694, 104 S.Ct. 2052, 80 L.Ed.2d 674.
{¶ 115} The state argues that even if trial counsel’s mitigation investigation was deficient, the court of appeals erred in concluding that Herring was prejudiced as a result of the deficiency without reweighing the evidence. The state claims that a reweighing of the evidence establishes that Herring was not prejudiced.
1. Legal standard
{¶ 116} In assessing prejudice, “the question is whether ‘there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different. A reasonable probability is a probability sufficient to undermine confidence in the outcome.’ ” State v. Williams, 99 Ohio St.3d 493, 2003-Ohio-4396, 794 N.E.2d 27, ¶ 163, quoting Strickland at 694. To assess that probability, we consider “the totality of the available mitigating evidence” and reweigh it “against the evidence in aggravation.” Williams v. Taylor at 397-398.
{¶ 117} Additional mitigating evidence that is “ ‘merely cumulative’ of that already presented” does not undermine the results of sentencing. Broom v. Mitchell, 441 F.3d 392, 410 (6th Cir.2006), quoting Clark, 425 F.3d at 286. Instead, “the new evidence * * * must differ in a substantial way — -in strength and subject matter — from the evidence actually presented at sentencing.” Hill v. Mitchell, 400 F.3d 308, 319 (6th Cir.2005); see Tibbetts v. Bradshaw, 633 F.3d 436, 444 (6th Cir.2011).
2. Evidence presented at the mitigation hearing
{¶ 118} Herring was sentenced to death for the course-of-conduct aggravating circumstance involving his intentional participation in three murders and two attempted murders. The evidence at trial showed that Herring and his accomplices killed these victims during a planned robbery of a Youngstown bar. In this court’s independent sentence review on direct appeal, we described the events that occurred at the Newport Inn and discussed evidence that singled out Herring’s culpability as the ringleader:
The robbers clearly coordinated their actions in advance. They discussed the robbery among themselves before going to the Newport Inn; they divided into two groups to cover both doors; they started shooting almost immediately. The coordination displayed here belies the notion *192that the killings were merely impulsive acts by individual members of the gang. That coordination supports the conclusion that the killings were integral to the robbery plan and that each of the robbers intended to kill the victims as part of that plan.
And that inference is especially strong with regard to Herring, because he was the evident ringleader. It was at his house that the robbers assembled, and he initiated the discussion of the robbery. Herring was the only robber prepared with a mask. He also obtained the guns (except Foose’s), and he decided who would carry which gun.
94 Ohio St.3d at 266, 762 N.E.2d 940.
{¶ 119} We also explained why Herring should receive the death penalty even though Antwan Jones, one of Herring’s codefendants, did not. Jones was convicted of three counts of aggravated murder and two counts of attempted aggravated murder, but death specifications against him were dismissed. Id. at 267. We stated:
[Sjince Herring was the leader, the state certainly had a stronger case against him than against Jones on the issue of intent. Moreover, the state had ballistics evidence pointing to Herring, not Jones, as the actual killer on Count One (even though the jury acquitted Herring of being the actual killer on that count). There was no such evidence against Jones, who was carrying a .357-caliber firearm rather than a 9 mm. The state thus had a rational basis to seek the death penalty for Herring and not for Jones.
Id. at 268.
{¶ 120} As for mitigation, trial counsel presented evidence that Herring’s accomplices did not receive the death penalty. Herring’s mother and his older sister also provided positive testimony about his loving relationship with his family and asked the jury to spare his life. In addition, trial counsel argued that the jury should consider Herring’s youth. Herring was 18 years and 8 months old at the time of the murders. Id. at 267.
3. Undiscovered evidence
{¶ 121} In contrast to the evidence produced during the mitigation hearing, the evidence that Herring’s counsel could have produced if they had conducted an adequate investigation presents a more comprehensive picture of Herring’s troubled background. Compelling mitigation evidence that was never presented to the jury indicated that (1) Herring’s father died violently in a drug dispute when Herring was young, (2) his mother used crack cocaine for most of Herring’s *193childhood, (3) his stepfather was addicted to drugs, (4) Herring began abusing drugs and alcohol at an early age and used them almost daily, (5) Herring dropped out of school before finishing the tenth grade, (6) his mother did not know whether Herring had ever graduated from high school, (7) Herring was a gang member for much of his life, and (8) Herring began selling drugs in his early teens.
{¶ 122} In addition, Dr. Brams could have presented information about Herring’s dysfunctional childhood. The jury would have learned that Herring “was raised in an environment in which * * * basically every parental figure, caregiver, family member, and associate was involved in illegal activities, significant drug and alcohol abuse, and to whom the consequences of violating the law evoked little anxiety.” The jury could have also heard that Herring was introduced to gangs and drug dealing through other family members and that Herring’s own mother asked him to purchase drugs for her. Thus, in contrast to the testimony presented at trial, substantial evidence showed that Herring had not been raised by a caring and nurturing family.
{¶ 123} Other evidence that was not presented to the jury revealed cognitive and learning-skills deficits. Herring had a full-scale IQ of 87. But other test results showed that Herring had a low score on visual-motor functioning and had the perceptual-learning skills of a ten-year-old. In addition, Dr. Brams diagnosed Herring with chronic alcohol abuse, cannabis abuse/dependence, polysubstance abuse/dependence, depressive disorder, personality disorder with narcissistic and antisocial features, and learning disabilities. Dr. Brams also stated that Herring’s IQ and achievement profiles, his history (which is suggestive of learning disabilities), and his chronic and early onset of substance abuse showed that he should have had a neuropsychological evaluation to determine whether he suffers from an organic brain impairment. Thus, additional information about Herring’s mental impairments could have been developed and presented to the jury if counsel had performed an adequate investigation.
4. State’s position
{¶ 124} The state argues that Herring was not prejudiced by trial counsel’s failure to ensure that a more complete mitigation investigation occurred. First, the state argues that the mitigation testimony of many of the family members would not have been effective because a majority of them stated that they did not believe that Herring committed the offenses. But this does not mean that the underlying information that these witnesses could have conveyed about Herring’s background was not believable.
{¶ 125} Second, the state argues that the mitigating evidence that trial counsel failed to discover would not have made a difference in the outcome of the sentencing proceedings if that evidence had been presented to the jury. The *194state cites Nields v. Bradshaw, 482 F.3d 442, 454 (6th Cir.2007), and Keith v. Mitchell, 455 F.3d 662, 670 (6th Cir.2006), in arguing that the Sixth Circuit has found that the failure to present similar background information about a capital defendant was not prejudicial.
{¶ 126} In Nields, trial defense counsel had failed to discover that
“ ‘Nields’s childhood home life was chaotic and neglectful,’ that ‘he was an expert and dedicated musician whose life was once very focused,’ that ‘he had several successful employment experiences and was a hard worker,’ * * * and that he ‘was a dependable kind-hearted friend and an extremely helpful, friendly person.’ ”
Id. at 454, quoting Nields’s brief. The court found that this additional information was “largely cumulative” of testimony at trial. Id. Unlike in this case, in Nields the additional evidence did not include gang involvement, drug and alcohol abuse, or possible mental problems. Also unlike in this case, the court in Nields found that the additional information was largely cumulative of evidence presented at trial.
{¶ 127} In Keith, postconviction affidavits showed that trial defense counsel had failed to interview family members, that friends and family had praised Keith’s abilities in high school football and his relationship with his daughter and nieces, and that a forensic and neuropsychological consultant had opined that Keith might suffer from a mild brain impairment. Id. at 670. Evidence was also presented that his mother was a drug addict, that he had been raised by his grandparents, that his grandmother was a convicted murderer, and that his father gambled and was known to run the streets. Id. In concluding that Keith was not prejudiced, the Sixth Circuit explained: “The additional mitigating evidence * * * does not demonstrate that Keith’s life had been so terrible that he was materially less culpable. In addition, * * * much of the so-called additional information was already given to the jury in the [presentence report], including descriptions of Keith’s family history and childhood circumstances.” Id.
{¶ 128} There are similarities between the additional evidence in Keith and that in the present case. Both cases involve a family history of neglect, drug abuse, and criminal misconduct, and both involve the possibility that the defendant suffered from brain impairment. But unlike in the present case, Keith’s jury had been provided a presentence report that described his family history and childhood circumstances.
{¶ 129} Moreover, the Sixth Circuit has found that defense counsel’s failure to conduct a full mitigation investigation was prejudicial in a number of different circumstances that are similar to the present case. See, e.g., Foust, 655 F.3d at *195539-546 (prejudice found when counsel had failed to obtain information about family chaos, parental abuse and neglect, squalor, incest, and sexual abuse in the family home and about good acts defendant performed in saving a baby from being shot in a drive-by shooting and convincing his sister to stop strip dancing and using drugs); Johnson, 544 F.3d at 606 (prejudice found when counsel had failed to discover evidence that defendant’s mother was a neglectful drug addict and had committed “atrocities” on him by giving him beer and Percocet when he was a child to stop crying, putting a cigarette out in his eye, and teaching him as an adolescent to prepare and sell crack cocaine); Jells, 538 F.3d at 500-501 (prejudice found when counsel had failed to discover evidence showing that defendant had significant learning disabilities, which led to aggressive behavioral responses, and that he experienced a sense of victimization due to his mother’s abusive relationships).
{¶ 130} Third, the state argues that the court of appeals ignored the fact that Herring’s and his family’s alcohol and drug abuse are entitled to little weight in mitigation. The additional mitigation that counsel failed to discover, however, involved much more than Herring’s alcoholism and drug abuse.
{¶ 131} Finally, the state argues that the trial judge’s opinion overruling Herring’s postconviction claim should be afforded great deference because the same judge presided over Herring’s trial and postconviction proceedings. See Williams, 529 U.S. at 396-397, 120 S.Ct. 1495, 146 L.Ed.2d 389; Schriro, 550 U.S. at 476, 127 S.Ct. 1933, 167 L.Ed.2d 836. That judge never reached the prejudice prong of Herring’s ineffectiveness claim, however, because he ruled that trial counsel’s mitigation investigation was not deficient.
5. Reweighing the aggravating circumstance and mitigating factors
{¶ 132} In this case, counsel presented some evidence at the mitigation hearing. The judge and jury heard almost nothing that would have humanized Herring or allowed them to gauge his moral culpability. They learned about Herring’s crimes, that his mother and sister loved him, that his accomplices did not receive the death penalty, that he was young, and almost nothing else.
{¶ 133} Had Herring’s counsel been effective, the judge and jury could have learned of the “kind of troubled history” that the United States Supreme Court has “declared relevant to assessing a defendant’s moral culpability.” Wiggins, 539 U.S. at 535, 123 S.Ct. 2527, 156 L.Ed.2d 471. They could have heard many specific details about (1) Herring’s dysfunctional childhood, (2) his family history of alcohol and drug abuse, (3) his gang involvement, (4) his mental-health problems, and (5) his possible brain impairment. See Penry v. Lynaugh, 492 U.S. 302, 319, 109 S.Ct. 2934, 106 L.Ed.2d 256 (1989), abrogated on other grounds, Atkins v. Virginia, 536 U.S. 304, 122 S.Ct. 2242, 153 L.Ed.2d 335 (2002), quoting California v. Brown, 479 U.S. 538, 545, 107 S.Ct. 837, 93 L.Ed.2d 934 *196(1987) (O’Connor, J., concurring) (“ ‘evidence about the defendant’s background and character is relevant because of the belief, long held by this society, that defendants who commit criminal acts that are attributable to a disadvantaged background * * * may be less culpable’ ”).
{¶ 134} On the other side of the ledger, we acknowledge that Herring’s course of conduct in killing three people and attempting to kill two more was horrific. “Powerful aggravating circumstances, however, do not preclude a finding of prejudice.” Foust, 655 F.3d at 546. Substantial mitigating factors existed in this case. The undiscovered and omitted evidence detailed above provided a compelling narrative that could have shifted the balance between the aggravating circumstance and the mitigating factors. Although we express no view on whether the aggravating circumstance outweighs the mitigating factors, we conclude that there is a reasonable probability that the penalty-phase outcome would have been different but for the errors of defense counsel.
YI. Conclusion
{¶ 135} We hold that trial counsel were deficient by failing to conduct a thorough and adequate investigation into Herring’s background before his mitigation hearing. We also hold that the court of appeals properly determined that counsel’s deficiency was prejudicial pursuant to Strickland. Accordingly, we affirm the judgment of the court of appeals vacating Herring’s death sentence and remanding this matter to the trial court for a new sentencing hearing, at which a new jury shall be impaneled to consider whether to impose the death penalty or a life sentence.
Judgment affirmed.
O’Connor, C.J., and French and O’Neill, JJ., concur. O’Donnell, Lanzinger, and Kennedy, JJ., dissent.