279 A.D.2d 770
N.Y. App. Div.
2001
Mugglin, J.

Cross appeals from a judgment of the Supreme Court (Malone, Jr., J.), entered December 3, 1999 in Albany County, which dismissed petitioners’ application, in a proceeding pursuant to CPLR article 78, to review determinations of respondents expelling petitioners Carl A. Basile and Daniel R. Papelino from respondent Albany College of Pharmacy, and awarding petitioner Michael Yu a failing grade in a required course.

Petitioners, fourth year students at respondent Albany College of Pharmacy (hereinafter the College), were charged by several of their respective professors with cheating on examinations given in various courses spanning two academic years. Following a hearing conducted by respondent Albany College of Pharmacy Student Honor Code Committee (hereinafter the Committee), petitioners were found guilty of cheating despite the absence of any evidence as to the specific means by which they allegedly cheated. The evidence in support of the charges consisted of (1) compilations by the various professors showing that petitioners gave the same incorrect answers to multiple choice examinations, (2) two anonymous notes, one of which *771claimed that two petitioners requested information concerning the contents of exams and the second questioning whether the same two petitioners were cheating, and (3) similar answers to questions which required calculations, although each petitioner utilized different calculations uncorrelated to the answer arrived at.

Petitioners commenced this CPLR article 78 proceeding to annul respondents’ determinations. Although the petition alleged certain procedural defects in the administrative hearing process, by stipulation and order petitioners waived all procedural defects underlying the determinations. Supreme Court, concluding that the appropriate standard of review was whether the determinations were rationally based, dismissed the petition. As a collateral matter, Supreme Court refused to consider an affidavit offered by the College since it was not a part of the administrative record underlying the determinations under review. Petitioners now appeal Supreme Court’s dismissal of the petition and confirmation of the determinations. Respondents cross-appeal from Supreme Court’s refusal to consider the affidavit.

As an initial matter, we observe that Supreme Court adopted the correct standard of review. It is well settled that judicial review of academic disciplinary matters between a private college and its students, where a hearing is not required by law, is whether the determination is arbitrary or capricious, not whether it is supported by substantial evidence (see, Matter of Rensselaer Socy. of Engrs. v Rensselaer Polytechnic Inst., 260 AD2d 992, 993; Matter of Christopher v Phillips, 160 AD2d 1165, 1167, lv denied 76 NY2d 706). “Rationality is what is reviewed under both the substantial evidence rule and the arbitrary and capricious standard [citation omitted]” (Matter of Pell v Board of Educ., 34 NY2d 222, 231). Unlike Supreme Court, however, we do not find that the Committee’s determinations that petitioners cheated on various examinations have a rational basis and we therefore reverse.

First, the Committee’s determinations were based solely on the statistical compilation. While these may give rise to a suspicion of cheating, suspicion alone will not suffice (see, Matter of Chiaino v Lomenzo, 26 AD2d 469, 473). An affidavit from an expert statistician, unrebutted by the College, establishes that the statistical case propounded by the professors is based upon false assumptions and therefore does not provide a rational basis to conclude that petitioners cheated. Specifically, the expert’s opinion points out that the statistics are valid only if the persons taking the examination had no knowledge of the *772subject matter (“randomness”) and had not studied together (“independence”).

Second, petitioners Carl A. Basile and Daniel R. Papelino were charged with cheating in nine courses and petitioner Michael Yu in seven courses. Basile was found guilty of cheating in six of the nine, Papelino in three of the nine and Yu in one of the seven. Since the same statistical methodology was used in every instance, there is no rational explanation which would support a finding of guilt in some courses but not in others.

Third, respondents claim that in addition to these statistics, there is evidence in the record of cheating by petitioners. A careful review of these allegations reveals that they are either hearsay anonymous notes or based on sheer speculation, neither of which will rationally support the determinations of the Committee (see, 300 Gramatan Ave. Assocs. v State Div. of Human Rights, 45 NY2d 176, 180). Moreover, it was irrational of the Committee to determine that it could rely solely on the inference of cheating raised by the statistical compilation, particularly when faced with proof that petitioners took these examinations in separate rooms and under the watchful eye of a proctor, who discerned no evidence of cheating.

As a final matter, Supreme Court correctly refused to consider the affidavit offered by the College as part of its answer to the CPLR article 78 proceeding to support the determination. Clearly, this affidavit was not part of the administrative record forming the basis for the administrative determinations (see, Matter of Levine v New York State Liq. Auth., 23 NY2d 863, 864).

Crew III, Rose and Lahtinen, JJ., concur; Cardona, P. J., not taking part. Ordered that the judgment is reversed, on the law, with costs, determinations annulled and petition granted.