JULIA SMITH GIBBONS, Circuit Judge.
Plaintiff-appellant Rose Dragon appeals the district court’s decision affirming the decision of Michael J. Astrue, Commissioner of Social Security, denying her supplemental security income under Title XVT of the Social Security Act. For the following reasons, we reverse the decision of the [456]*456district court and remand to the administrative law judge for an immediate award of benefits.
I.
Rose Dragon (“Dragon”) was born on February 29, 1980. She is married and lives with her husband and daughter. She and her husband share the household tasks like grocery shopping, caring for her daughter, laundry, and dishes. She has a high school education (completed through special education classes) and has attempted to work a number of jobs, including dishwasher, bagger at a grocery store, and fast-food worker. Her employment was always short-lived because, as she testified, she “got frustrated,” she “wasn’t fast enough,” or “if people made fun of [her], [she] would get discouraged.” In June 2004, Dragon applied for supplemental security income (“SSI”) under Title XVI of the Social Security Act alleging disability due to speech problems, learning difficulties, and asthma onset as of April 1, 2004. Upon denial of her claims she requested a hearing before an administrative law judge (“ALJ”).
The ALJ reviewed records from Dragon’s schools from 1994 to 1998. These records note that, in 1992 at the age of 12, Dragon took the Wechsler Intelligence Scale for Children-Third Edition (“WISC-III”), and received a verbal scale I.Q. score of 65, a performance scale I.Q. score of 72, and a full scale I.Q. score of 66. The same records indicate that Dragon had a “mild articulation delay” and that “Rosanna qualifie[d] for a program for developmentally handicapped and for speech therapy.” The school placed Dragon on an individual education plan for the '93 — '94 and '94-'95 school years. Dragon never passed any of the ninth grade proficiency tests and was exempted because she “[did] not have the necessary skills to pass.” Dragon was considered a 1998 graduate after she completed her requirements during the summer. A 1998 assessment stated that Dragon was fairly successful in her classes, that she took a regular curriculum with study skills for support, and that she received work-study credits. Post-graduation, the record,indicated that the “team [felt] that Rose [was] developing skills needed to live independently with some assistance.”
On July 23, 2004, clinical psychologist Dr. Paul Deardorff performed a consultative psychological evaluation of Dragon. Dr. Deardorff administered the Wechsler Adult Intelligence Scale-Third Edition (“WAIS-III”). Dragon received a verbal I.Q. score of 58, a performance scale I.Q. score of 51, and a full scale I.Q. score of 50, indicating that her current functioning was in the “moderately retarded range of intelligence or at least well less than the percentile for [her] age group.” In his report, Dr. Deardorff stated, “[Dragon’s] tested I.Q. is lower than would be expected on the basis of clinical presentation. Emotional and motivational factors may have interfered with performance on this measure as [she] appears to be of borderline intelligence.” In his multiaxial classification of Dragon, on Axis I he listed her clinical disorders as anxiety disorder and reading disorder; on Axis II rather than listing a level of mental retardation Dr. Deardorff classified Dragon as of “borderline intellectual functioning”; Dr. Deardorff listed no general medical conditions under Axis III; on Axis IV Dr. Deardorff listed Dragon’s psychosocial and environmental problems as occupational problems, problems interacting with social environment, problems with primary support group, and educational problems; and on Axis V he assessed a Global Assessment of Functioning (“GAF”) score of 55. In reaching Dragon’s GAF score, Dr. Deardorff noted Dragon’s anxiety and irrita[457]*457tion. He observed that she “has few friends but sees them on a daily basis ... [,] cares for her daughter and assists with household chores ... [,] plays video games, listens to the radio[,] and likes to shop.” A GAF score of 55 indicates moderate symtomatology and difficulties in social, occupational, or school functioning.
Regarding Dragon’s work-related mental abilities, Dr. Deardorff observed:
1. The claimant’s mental ability to relate to others, including fellow workers and supervisors is moderately to seriously impaired by her emotional difficulties. She was somewhat uncomfortable dealing with this examiner and she stated that she believes others ridicule her. She very likely would have difficulty relating adequately to others in completing simple repetitive tasks.
2. The claimant’s mental ability to understand, remember, and follow instructions is moderately impaired by her emotional difficulties. She would have no difficulty understand [sic] simple instructions but her short term memory skills are weak and she may have difficulty retaining them. Further, her pace may be slowed by her depressive symptomatology.
3. The claimant’s mental ability to maintain attention, concentration, persistence, and pace is moderately impaired by her emotional difficulties. Her attention and concentration skills were not strong during this evaluation and may deteriorate over extended time periods, slowing her performance in completing simple repetitive tasks.
4. The claimant’s mental ability to withstand the stress and pressure associated with day-to-day work activities is moderately impaired by her emotional difficulties. Such stress may result in increased anxiety and decreased attention and concentration skills, and might also result in increased suspiciousness and misinterpretation of the behavior and statements of those around her.
5. Should she be granted disability compensation, she would have difficulty in managing funds prudently as her arithmetic skills are limited.
On August 4, 2004, Dr. Douglas Pawlarczyk, a state agency psychologist, filled out a psychiatric review technique form (“PRTF”) reviewing Dragon’s record, which included her school records, Dragon’s contact with the Agency, and Dr. Deardorffs examination. The PRTF largely consists of checked boxes indicating that Dr. Pawlarczyk found that medically determinable impairments were present but did not precisely satisfy the diagnostic criteria for mental retardation and anxiety-related disorders. He further checked that she had mild restriction of activities of daily living; moderate difficulties in maintaining social function and maintaining concentration, persistence, or pace; and no evidence of related episodes of decompensation of extended duration. The last page of the PRTF contained a functional capacity assessment, the only portion of the PRTF actually written by Dr. Pawlarczyk. This largely reiterated Dr. Deardorffs findings of Dragon’s functional capacity. Dr. Pawlarczyk did note, in particular, that “[Dr. Deardorff] felt that [Dragon’s I.Q. scores] may be an underestimate as [Dragon] clinically presented as more in the [borderline intellectual functioning] range. Those school age score [sic] are more likely an [sic] true estimate of her intellectual ability.” Alice, Psy.D., a state agency psychologist, stamped her approval on the PRTF on September 23, 2004.
[458]*458Dragon, her father, and a vocational expert (“VE”) testified at a hearing on April 17, 2006, before the ALJ. Dragon testified to her home life, her work experiences, and her general disabilities. The VE testified regarding the skill level of Dragon’s only durationally significant job at the Old Country Buffet: the dishwasher job rated as medium work and unskilled, the busser job as medium work and unskilled, and the attendant job as light work and unskilled. The VE noted that, similar to the dish-washing job, there are cleaning jobs at a medium level and a light level that are appropriate for an individual with a marginal education and who is able to perform only simple, routine tasks; can understand, remember, and carry out only short and simple instructions; can interact no more than occasionally with the general public; can make only simple work-related decisions; and cannot work at a rapid product rate pace. But upon examination by Dragon’s counsel, the VE revealed that Dragon’s problems interacting with others, keeping up with tasks, and inability to handle certain stresses might interfere with this kind of work.
On May 23, 2006, the ALJ issued an unfavorable decision in Dragon’s case, ultimately concluding that Dragon was not disabled, was able to perform past work, and thus was not entitled to SSI. In reaching this determination, the ALJ conducted the five-step evaluation pursuant to 20 C.F.R. § 416.920. After determining that Dragon had not been engaged in substantial gainful activity since April 1, 2004, the ALJ recognized that Dragon did indeed have some severe impairments. Among these severe impairments was borderline intellectual functioning, which the ALJ affirmatively found to exist. The ALJ determined, however, that Dragon failed on the third, fourth, and fifth steps of the analysis.
As to the third step, the ALJ addressed Dragon’s severe impairments — specifically borderline intellectual functioning, a mild articulation delay, an anxiety disorder, and a reading disorder — but found that “the severity of [those] impairments, either singly or in combination, [did] not meet or equal the requirements of any listed impairment in Appendix 1 to Subpart P of 404, including Sections 12.05 and 12.06.” The ALJ again noted that Dragon is “of borderline intelligence and is not mentally retarded.” The ALJ specifically rejected a portion of Dr. Deardorff s analysis: “The I.Q. scores obtained by Dr. Deardorff were felt to be underestimates of the claimant’s intellectual functioning and therefore are not valid.” The ALJ referenced I.Q. scores in the mid-60s obtained while Dragon was a child but failed to consider the scores because the testing “is not actually contained in the file to assess whether the results were considered to be valid.” The ALJ then went on to note that Dragon’s high school education was not consistent with mental retardation. The ALJ concluded his third step analysis as follows: “There is very little to no substantive indication that the claimant has a history of deficits in adaptive functioning initially manifested during the developmental period to the attainment of age 22.”
With regard to the remaining two steps, the ALJ set forth the following residual functional capacity (“RFC”):
[D]ue to her severe impairments, the [ALJ] finds that [Dragon] can perform only simple, routine tasks and is able to understand, remember, and carry out only short and simple instructions. She cannot interact with the general public more than occasionally. She cannot work at a rapid production-rate pace. She is able to make only simple work-related decisions. Any job should not require the ability to make change, or [459]*459reading or writing skills above the third grade level. The job should not require more than very simple math. [Dragon] has no exertional limitations.
Relying on the VE’s assessment, the ALJ concluded that Dragon “retains the ability to perform her past relevant dishwasher work and other cleaner jobs despite her impairments,” thus failing to establish a disability under the fourth and fifth steps of the analysis. The ALJ ultimately denied Dragon’s claim for SSI.
Dragon filed suit in the United States District Court for the Southern District of Ohio, seeking judicial review of the ALJ’s decision. On January 18, 2009, a magistrate judge issued a Report and Recommendation (“R&R”), recommending that the case be reversed and remanded for an immediate award of benefits. The Commissioner filed an objection to the R&R, and Dragon filed a response. The district court declined to adopt the magistrate’s R&R and affirmed the ALJ’s decision. Dragon now appeals on the ground that substantial evidence did not support the ALJ’s finding that she was not disabled and therefore not entitled to SSI.
II.
We review the district court’s conclusion in a social security case de novo. Valley v. Comm’r of Soc. Sec., 427 F.3d 388, 390 (6th Cir.2005). We review the underlying ALJ’s findings to determine whether they are supported by substantial evidence. Id. (citing 42 U.S.C. § 405(g)). “Substantial evidence is more than a scintilla of evidence but less than a preponderance; it is such relevant evidence as a reasonable mind might accept as adequate to support a conclusion.” Cutlip v. Sec’y of Health & Human Servs., 25 F.3d 284, 286 (6th Cir.1994). If there is such support, those findings are conclusive, and this court may not reverse even if it would have arrived at a different conclusion. Valley, 427 F.3d at 391.
III.
Under the Social Security Act, “disability” is defined as the inability “to engage in any substantial gainful activity by reason of any medically determinable physical or mental impairment which can be expected to result in death or which has lasted or can be expected to last for a continuous period of not less than 12 months.” 42 U.S.C. §§ 423(d)(1)(A), 1382c(a)(3)(A). Dragon has the burden of establishing that she meets this definition. 42 U.S.C. §§ 423(d)(5)(A). The evaluation is a five-step, sequential process, and if the ALJ finds disability or lack of disability at any step, the determination is made without proceeding to the next step. 20 C.F.R. § 416.920(4). The steps are as follows:
(i) At the first step, we consider your work activity, if any. If you are doing substantial gainful activity, we will find that you are not disabled....
(ii) At the second step, we consider the medical severity of your impairment(s). If you do not have a severe medically determinable physical impairment that meets the duration requirement ..., or a combination of impairments that is severe and meets the duration requirement, we will find that you are not disabled ....
(iii) At the third step, we also consider the medical severity of your impairments). If you have an impairment(s) that meets or equals one of our listings in appendix 1 to subpart P of Part 404 of this chapter and meets the duration requirement, we will find that you are disabled....
(iv) At the fourth step, we consider our assessment of your residual functional capacity and your past relevant work. [460]*460If you can still do your past relevant work, we will find that you are not disabled ....
(v) At the fifth and last step, we consider our assessment of your residual functional capacity and your age, education, and work experience to see if you can make an adjustment to other work. If you can make an adjustment to other work, we will find you are not disabled. If you cannot make an adjustment to other work, we will find that you are disabled.
Id. § 416.920(a)(4)(i)-(v).
According to the ALJ, Dragon met her burden in steps 1 and 2, but failed on steps 3, 4, and 5. We disagree. We conclude that the ALJ’s decision to reject the evidence in the record supporting Dragon’s claims of impairment and to affirmatively find that Dragon was of borderline intellectual functioning as opposed to mentally retarded was not supported by substantial evidence. Our discussion will elaborate on why the ALJ erred in finding that Dragon failed to develop her disability on steps 3, 4, and 5.
A.
The third step requires an examination of the medical severity of Dragon’s impairments and a determination of whether they meet one of the listings found in appendix 1. Dragon argues that she meets the listing for mental retardation, specifically Listing 12.05C, which has three requirements: (1) “significantly subaverage general intellectual functioning with deficits in adaptive functioning initially manifested during the developmental period; i.e., the evidence demonstrates or supports onset of the impairment before age 22”; (2) “[a] valid verbal, performance, or full scale IQ of 60 through 70”; and (3) “a physical or other mental impairment imposing an additional and significant work-related limitation of function.” 20 C.F.R. Pt. 404, Subpt. P, App. 1, § 12.05C.1 The ALJ determined that Dragon failed to meet this listing because of his affirmative finding that Dragon “is of borderline intelligence and not mentally retarded.” The ALJ ignored the wealth of evidence in support of Dragon’s claims and affirmatively found otherwise.
Regarding the first requirement, the ALJ stated that “[tjhere is very little to no substantive indication that [Dragon] has a history of deficits in adaptive functioning initially manifested during the developmental period prior to the attainment of age 22.” “Adaptive functioning includes a claimant’s effectiveness in areas such as social skills, communication, and daily living skills”. West v. Comm’r Soc. Sec., 240 Fed.Appx. 692, 698 (6th Cir.2007) (citing Heller v. Doe by Doe, 509 U.S. 312, 329, 113 S.Ct. 2637, 125 L.Ed.2d 257 (1993)). Dragon, however, did present significant evidence to the contrary, including a set of I.Q. scores obtained in 1992 when she was 12-years old. This testing resulted in a verbal I.Q. score of 65, a performance I.Q. score of 72, and a full scale I.Q. score of [461]*46166. The ALJ disregarded these scores because “this testing was 14 years ago and occurred while [Dragon] was a child.”2 Disregarding the scores on this ground misconstrues the relevance of the scores; indeed, an older score was relevant to establish the manifestation of Dragon’s impairment before the age of 22. Cf. West, 240 Fed.Appx. at 698 (“While the claimant may use a qualifying IQ score before the age of 22 to demonstrate that his subaverage intellectual functioning initially manifested during his developmental period, ... a claimant is by no means required to produce an IQ score obtained prior to age 22.”).
The ALJ also disregarded these scores because they were “only referenced in the file and [were] not actually contained in the file to assess whether the results were considered to be valid.” Nothing in the school records suggests that the scores were invalid, whether or not the actual testing was contained in the file. The scores and school records indicating that Dragon was to proceed in a separate educational program suggest the manifestation of Dragon’s impairment before the age of 22. Moreover, the records indicate that she had a “mild articulation delay,” qualified for a program “for developmentally handicapped and for speech therapy,” and did not have the skills to pass her ninth grade proficiency tests. The I.Q. scores were consistent with the evidence in the school records and were valid to prove one of the necessary elements of her disability claim — that her deficits in adaptive functioning manifested prior to the age of 22. For those reasons, we find that the ALJ lacked substantial evidence to determine that Dragon’s childhood I.Q. scores were invalid and lacked substantial evidence to conclude that Dragon had failed to establish “significantly subaverage general intellectual functioning with deficits in adaptive functioning initially manifested during the developmental period.”
The second requirement under Listing 12.05C is “[a] valid verbal, performance, or full scale I.Q. of 60 through 70.” To establish this, Dragon submitted the set of I.Q. scores obtained by Dr. Deardorff, in which Dragon scored a verbal scale I.Q. of 58, performance scale I.Q. of 51, and full scale I.Q. of 50 on the WAIS-III. The ALJ concluded that these scores were not valid because “[t]he I.Q. scores obtained by Dr. Deardorff were felt to be underestimates of [Dragon’s] intellectual functioning.” This conclusion was based on Dr. Deardorffs Axis II categorization of “borderline intellectual functioning” and Dr. Deardorffs statement that Dragon’s “tested I.Q. [was] lower than would be expected on the basis of clinical presentation,” but that “[e]motional and motivational factors may have interfered with performance on this measure as [Dragon] appeared] to be of borderline intelligence.” An affirmative finding of borderline intellectual functioning is usually associated with I.Q. scores of at least 70.
The ALJ ignored Dr. Deardorff s earlier observation that Dragon “did not appear [462]*462to exaggerate or minimize her difficulties ... [and] was adequately motivated.” In spite of his “borderline intellectual functioning” classification and high expectations, Dr. Deardorff acknowledged that Dragon’s I.Q. scores suggested that her performance falls “moderately below the average range” on tasks calling for the comprehension of words, abstract verbal reasoning abilities, a fund of information, or common-sense reasoning abilities. He further observed that she falls “well below average limits” in her perceptual-organizational or non-verbal problem-solving abilities, arithmetic reasoning skills, and short-term memory skills. Dr. Deardorff also discussed his administration of the Wechsler Memory Scale-Form III. Dragon’s memory abilities and attention and concentration skills fall within the moderately retarded range — “at well less than the 1st percentile for her age group” — and her reading abilities fall in the mildly retarded range — “at about the 1st percentile for her age group.” Dragon’s “[w]ord recognition abilities fall at the third grade level but her reading comprehension abilities are very likely not as well developed as her meager word recognition abilities.” Dr. Deardorff also assessed Dragon with a GAF of 55, indicating moderate symptomatology.
Notwitstanding the general consistency of Dragon’s test results, the ALJ placed great emphasis on the statement that Dragon’s tested I.Q. was lower than expected and found that Dr. Deardorffs statement on this point completely invalidated the I.Q. scores. This conclusion is not supported by substantial evidence. Dr. Deardorff apparently did expect Dragon to perform better on the I.Q. test based on her presentation. Yet, in order to actually test as of borderline intelligence, Dragon would have had to achieve an I.Q. score of 70, twenty points higher than she actually scored. Moreover, Dr. Pawlarczyk surmised that Dragon’s “school age score [sic] are more likely an [sic] true estimate of her intellectual ability.” Dragon’s school age scores placed her below the range of borderline intellectual functioning.
The remainder of Dr. Deardorffs report details Dragon’s substantial limitations, including poor attention, concentration, arithmetic, reading, and reasoning skills. The ALJ may choose to disregard I.Q. scores that would normally lead to a finding of disability when those scores were undermined by a doctor’s full evaluation. See Daniels v. Comm’r of Soc. Sec., 70 Fed.Appx. 868, 869, 872 (6th Cir.2003) (finding that substantial evidence supported the ALJ’s finding that claimant was not mentally retarded although claimant had a performance I.Q. of 67 because the examining doctor also found claimant to be within the “borderline range of intelligence” and clinically “appeared to function in the dull-normal range of intelligence”). But that was not the case here. The ALJ disregarded Dr. Deardorffs full evaluation, which served to reinforce the I.Q. scores, rather than undermine them.
Further in support of his decision to reject the I.Q. scores, the ALJ stated, “[T]he fact that [Dragon] was able to graduate from high school, albeit with an individualized education plan, is not consistent with mental retardation. She is married and has a six year old daughter who she appears to do a good job taking care of.” The regulations permit the ALJ to question the validity of test results in conjunction with other factors. “In assessing the validity of a claimant’s I.Q., ‘[finformation from both medical and nonmedical sources may be used to obtain detailed descriptions of the individual’s activities of daily living; social functioning; concentration, persistence and pace; or ability to tolerate [463]*463increased medical demands (stress).’ ” Brown v. Sec’y of Health & Human Servs., 948 F.2d 268, 269 (6th Cir.1991) (quoting 20 C.F.R. Pt. 404, Subpt. P, App. 1, § 12.00D). Nonetheless, the Sixth Circuit has found that certain activities are not necessarily “inconsistent with a valid test I.Q. of 68” — use of public transit, making change at a grocery‘story, doing laundiy (but with strong body odor), cleaning, limited reading comprehension (equivalent to a third-grade education), or the ability to keep records of work activity. Id. at 270. But see West, 240 Fed.Appx. at 698 (citing claimant’s ability “to drive a garbage truck on a part-time basis, to care for his daily needs, to pay bills, to shop for groceries, to interact with friends and families, and to engage in numerous other daily activities” as evidence that claimant was not deficient in adaptive functioning).
In this case, the ALJ reasoned that, because Dragon was a high school graduate and a mother, her I.Q. scores were invalid. Given the record here, these things cannot be held against her. See 20 C.F.R. § 416.920(d) (“If you have an impairments) which ... is listed in Appendix 1, we will find you disabled without considering your age, education, and work experience.”). She graduated from high school on an individual education plan and never passed her ninth grade proficiency tests. Although she has a good relationship with her daughter, all the while she has been living with another parental figure — first her daughter’s father and now her new husband. Because Dragon’s activities are not inconsistent with her tested I.Q. scores and Dragon’s performance during Dr. Deardorff s evaluation was not inconsistent with her tested I.Q. scores, the ALJ’s decision to invalidate the I.Q. scores was not supported by substantial evidence.
The third requirement for meeting Listing 12.05C is a physical or other mental impairment imposing an additional and significant work-related limitation of function. The ALJ appeared to reject the evidence presented by Dragon on this point by relying on the findings of Dr. Pawlarcyzk. The ALJ stated, “With specific regard to the ... criteria pertaining to the claimant’s mental impairments, the [ALJ] agrees with the findings of Dr. Pawlarcyzk that [Dragon] has a mild restriction of activities of daily living, moderate difficulties in maintaining concentration, persistence or pace, and no episodes of decompensation.”3 This reliance appears to have affected the ALJ’s finding that the severity of Dragon’s impairments did not meet Listing 12.05C as well as the ALJ’s finding of her RFC, which affects the fourth step of the analysis.
Dr. Deardorff performed a consultative psychological evaluation of Dragon on July 13, 2004. On August 4, 2004, Dr. Pawlarczyk performed a psychiatric review of the record (the PRTF), which included Dragon’s school records and Dr. Deardorffs evaluation. In weighing medical opinions, generally, a treating source is to be given more weight than an examining source and an examining source (Dr. Deardorff) more weight than a non-examining source (Dr. Pawlarcyzk). 20 C.F.R. § 404.1527(d)(1)-(2).4 This is not a bright-line rule. See [464]*464Lyons v. Soc. Sec. Admin., 19 Fed.Appx. 294, 302 (6th Cir.2001) (“A non-examining physician’s opinion may be accepted over that of an examining physician when the non-examining physician clearly states the reasons that his opinions differ from those of the examining physicians.”). Whether an opinion is consistent with the record as a whole or supported by relevant evidence will also factor into the evaluation of medical opinions. 20 C.F.R. § 404.1527(d)(3)-(4). The weight of non-examining sources will be given weight dependant on the “degree to which they provide supporting explanations for their opinions.” 20 C.F.R. § 404.1527(d)(3).
In this case, Dr. Pawlarcyzk’s evaluation consisted almost entirely of checked boxes and a regurgitation of some of Dr. Deardorffs findings. He provided little explanation or support for his conclusion that Dragon did not satisfy the diagnostic criteria of mental retardation and anxiety-related disorders. Dr. Deardorffs analysis, however, was quite detailed and supported not only by his own evaluation, but also by the rest of the record. Dragon was able to graduate from high school but only with an individual education plan and regardless of her inability to complete some basic requirements. Additionally, Dragon has proved that she is not unemployed for lack of trying. Prior to applying for SSI, Dragon has attempted many jobs for short time periods but has been unable to maintain consistent employment. Dragon testified that she had difficulty keeping up, that she was made fun of, that she was slower than others, and that she would get discouraged and frustrated at her various jobs. Dr. Deardorffs analysis confirmed her difficulties in relating to others, remembering instructions, keeping pace, maintaining attention, and withstanding stress and pressure. He determined that she suffered from anxiety disorder, reading disorder, borderline intellectual functioning, occupational problems, problems interacting with social environment, problems with primary support group, and educational problems. His assessment was consistent with a finding of mental retardation under 12.05C, which requires a finding of “a physical or other mental impairment imposing an additional and significant work-related limitation of function.”5 Substantial evidence did not support the ALJ’s decision to disregard Dragon’s examining doctor’s evaluation in favor of a form with little substantive information. Dr. Deardorffs more complete evaluation demonstrated that Dragon met her burden of establishing that she met Listing 12.05C.
For the foregoing reasons, we conclude that the ALJ’s determination that Dragon failed to meet a listing in Appendix 1 was not supported by substantial evidence. The regulations suggest that the analysis may end here: “If we can find that [Dragon is] disabled or not disabled at a step, we make our determination or decision and we do not go on to the next step.” [465]*46520 C.F.R. § 416.920(a)(4). Nevertheless, we will continue with our evaluation of the ALJ’s decision to fully conclude that it was not supported by substantial evidence.
B.
If Dragon had failed to demonstrate her disability under the third step, the court would then move to the fourth step — an assessment of Dragon’s RFC and her past relevant work. In reaching a claimant’s RFC, the ALJ is to assess the claimant’s physical abilities, mental abilities, other abilities affected by impairments, and total limiting effects. 20 C.F.R. § 404.1545(b)-(e). The RFC is based on medical evidence but also statements made by the claimant on what she can do. Additionally, the ALJ will “consider descriptions and observations of [the claimant’s] limitations from [her] impairment ... provided by [the claimant], [the claimant’s family], neighbors, friends, or other persons.” Id. § 404.1545(a)(3). Dragon argues that, for the same reasons discussed above, the ALJ erred in relying on Dr. Pawlarcyzk’s evaluation alone. We agree.
If the ALJ had given any credence to Dr. Deardorffs evaluation, the RFC would have looked quite different. The ALJ’s RFC stated that Dragon “cannot interact with the general public more than occasionally,” while Dr. Deardorffs more thorough analysis added that Dragon’s “mental ability to relate to others, including fellow workers and supervisors is moderately to seriously impaired by her emotional difficulties .... She very likely would have difficulty relating adequately to others in completing simple repetitive tasks.” Consistent with Dr. Deardorff, the ALJ found that Dragon “can perform only simple, routine tasks and is able to understand, remember, and carry out only short and simple instructions.” The ALJ, however, ignored Dr. Deardorffs statement that Dragon’s “pace may be slowed by her depressive symptomatology.” The ALJ determined that Dragon “cannot work at a rapid production-rate pace” and “is able to make only simple work-related decisions.” Dr. Deardorff viewed Dragon as even more limited, noting that Dragon’s “attention and concentration skills were not strong during this evaluation and may deteriorate over extended time periods, slowing her performance in completing simple repetitive tasks.” Furthermore, the ALJ completely ignored Dr. Deardorffs assessment that Dragon’s mental ability to withstand stress and pressure of day-to-day work activities is moderately impaired by her emotional difficulties. He noted, “Such stress may result in increased anxiety and decreased attention and concentration skills, and might also result in creased suspiciousness and misinterpretation of the behavior and statements of those around her.” The ALJ and Dr. Deardorff agreed that Dragon’s reading and arithmetic skills are limited. We conclude that the ALJ’s RFC was not supported by substantial evidence.
Additionally, the ALJ’s incomplete RFC was insufficient to elicit apt testimony from the VE regarding Dragon’s ability to perform her past relevant work. The VE testified that Dragon’s previous work has ranged from medium to light work, all unskilled. The VE also testified that the education described by the ALJ was not a functionally limited education, but rather Dragon’s education would be considered “marginal.” Given the ALJ’s assessment of Dragon’s RFC, the VE testified that Dragon could perform her past work as a dishwasher. Based on this testimony, the ALJ remained confident that Dragon would be able to work — “despite her history of academic problems and some unsuccessful work attempts.” The [466]*466ALJ understood the VE’s testimony to say that “despite the[] limitations stemming from [Dragon’s] severe impairments, [Dragon] could perform her composite job as a dishwasher, both as actually performed and as customarily performed in the national economy.” We conclude that this is simply not the case. Had the VE been presented with a more accurate picture of Dragon’s limitations, the VE’s testimony could have easily differed.
Furthermore, Dragon has clearly put forth the effort to find work; she attempted and failed to work as a dishwasher at Old Country Buffet. In assessing a claimant’s RFC, the ALJ will consider her statements about what she can do. 20 C.F.R. § 404.1545(a)(3). And in assessing a claimant’s ability to perform past relevant work, the claimant is to provide information about work done in the past. Id. § 404.1560(b)(2). Then the ALJ may look to others who may know about the claimant’s work as well as to vocational experts. Id. Di'agon provided information regarding her past relevant work and testified that she quit her job at Old Country Buffet for a number of reasons:
[T]hey told me if I didn’t quit there, I would have got fired because they found out that I was pregnant and then I couldn’t keep up and like I was sick a lot and they told me they were going to fire me and they said I was also too slow.
Dragon also testified as to why she believes she is unable to work in general:
I’m unable because any job I tried in the past, I couldn’t keep up because I’m slower than others and like say if people made fun of me, I would get discouraged and that holds me like I can’t do it and I feel like, basically, I feel I’m too dumb to do it, like any kind of job.
The ALJ found Dragon’s testimony “credible to the extent that such testimony is consistent with the above [RFC] assessment.6 To the extent that any such testimony was reflective of a greater degree of limitations, such testimony was not found to be credible due to a lack of objective medical support.” This reasoning is circular. The ALJ found Dragon’s testimony credible to the extent it confirmed the ALJ’s own conclusion. Thus, it appears the ALJ afforded Dragon’s testimony no weight at all because the ALJ had come to its conclusion independent of the information from her testimony.
But Dragon’s testimony was consistent, both internally and with the record provided to the ALJ, especially Dr. Deardorffs analysis. Consistency tends to support credibility, see Rogers v. Comm’n of Soc. Sec., 486 F.3d 234, 248 (6th Cir.2007), and Dragon consistently presented that she was unable to keep up at her previous jobs, including her job as a dishwasher. Dr. Deardorff noted that Dragon has held “a lot” of jobs, which she has left due to relocation, ridicule, and inability. Dragon has demonstrated that she wanted to work and persistently attempted to find work. She has been unable to do so, and to conclude that she is able to perform previous work at which she earnestly failed is contrary to the record. Thus, we conclude that the ALJ’s RFC and the determination that Dragon was able to perform her past relevant work given that RFC were not supported by substantial evidence.
C.
The fifth step asks whether Dragon is capable of adjusting to other work. Regarding this step, the ALJ understood the [467]*467VE’s testimony to suggest that Dragon “could perform the jobs of unskilled medium cleaner ... and unskilled light cleaner ... despite the above limitations.” Indeed, when presented with the ALJ’s perception of Dragon’s RFC and her education and past work, the VE testified that Dragon would also be able to perform other unskilled work such as light cleaning jobs including cleaning hotel rooms. For the reasons provided above, we find that the ALJ’s assessment of Dragon’s RFC lacked the support of substantial evidence. We conclude, thus, that the VE’s testimony is not supported by the record as the VE was not presented with the full assessment of Dragon’s abilities.
IV.
In this case, we find not only that the ALJ’s decision lacked a foundation in substantial evidence, but also that the evidence of Dragon’s disability is so substantial that she is entitled to an immediate award of benefits. Dragon has established significantly subaverage general intellectual functioning with deficits in adaptive functioning before the age of 22, a valid I.Q. score between 60 and 70, and a physical or other mental impairment imposing an additional and significant work-related limitation of function. See 20 C.F.R. Pt. 404, Subpt. P, App. 1, § 12.05C. Furthermore, we find that substantial evidence did not support the ALJ’s conclusions that Dragon was of borderline intellectual functioning, able to perform past relevant work, and able to perform other work. In this case, we find that all factual issues have been resolved and that the record adequately establishes Dragon’s entitlement to benefits. See Faucher v. Sec’y of Health & Human Servs., 17 F.3d 171, 176 (6th Cir.1994).
For the foregoing reasons, we reverse the decision of the district court and remand to the ALJ for an immediate award of benefits.