delivered the opinion of the court:
On June 2,1971, defendant Elson Boose was indicted for the murder of Robert Jefferson, a guard at the Rlinois Industrial School for Boys where the then 15-year-old defendant was incarcerated for a juvenile offense. On February 22, 1972, defendant pleaded guilty to one count of murder and was sentenced to a term of 20 to 40 years in the Department of Corrections. On November 10, 1975, the conviction was reversed and remanded for entry of a new plea as the trial court had abused its discretion in requiring defendant to be shackled during a competency hearing. (People v. Boose (1975), 33 Ill. App. 3d 250, 337 N.E.2d 338.) On April 5,1977, that decision was affirmed. People v. Boose (1977), 66 Ill. 2d 261, 362 N.E.2d 303.
Upon remand, defendant entered a new plea of not guilty and was tried by jury in the circuit court of La Salle County. Defendant was found *459guilty of murder and again was sentenced to a term of 20 to 40 years in the Department of Corrections. Defendant now appeals this conviction, contending that the trial court erred in not declaring a mistrial when a defense witness disclosed a conversation with defendant at Stateville Penitentiary concerning defendant’s prior appeal and abused its discretion in ordering an excessive sentence considering his rehabilitative success.
At defendant’s trial, an important defense witness was Bennie Singleton. Singleton hgd been an inmate at the Illinois Industrial School for Boys at the time of the homicide and was later incarcerated at Stateville Penitentiary for an adult conviction. The thrust of Singleton’s testimony was that defendant had not committed the charged offense by stabbing the victim or by plotting with other inmates to do so. During cross-examination, the following exchange transpired:
“Q ¡Where are you at now?
A: I’m at Stateville.
Q: Did you ever see Elson Boose at Stateville?
A: Yes I did.
Q: Did you ever have a conversation with him?
A: Yes I have.
Q: Did you ever talk about this incident with him?
A: When I first seen him, you know, it was about the Appellate Court, something about his appeal or something. * *
At this point, defense counsel asked to be heard out of the presence of the jury and moved for mistrial on the ground that defendant was irreparably prejudiced by the jury’s awareness of his incarceration at Stateville and participation in an appeal. The People responded that they were attempting to impeach Singleton’s credibility by demonstrating that he and defendant discussed the case at Stateville and that Singleton had completely reviewed the case file. The trial court considered “probative value against the prejudicial value,” denied the motion, and instructed the People to avoid the line of questioning. The jury was then recalled and no further reference to the testimony was made at any point during the trial.
The situation at bar differs from more familiar contexts in which indicia of prior criminality are assigned as the basis of prejudicial error. The People made no attempt to introduce evidence of criminality indicating motive, intent, identity, absence of mistake, modus operandi, or false alibi (see, e.g., People v. Romero (1977), 66 Ill. 2d 325, 362 N.E.2d 288), or to impeach defendant’s credibility (e.g., People v. Montgomery (1971), 47 Ill. 2d 510, 268 N.E.2d 695). Second, the prior criminality involved was not that of a prior conviction but stemmed rather from defendant’s initial conviction in this cause. Third, the People’s attempt to impeach the defense witness created the possibility of the jury being *460influenced by that conviction, rather than, or in addition to, evaluating Singleton’s credibility.
Neither the parties litigant nor our research have revealed a precise benchmark of prejudicial error in this situation. Under these circumstances, we feel we must consider two standards in determining whether reversible error was here committed. We shall first consider the principles governing new trials and then consider those principles governing references to prior criminality in general.
On a new trial of a case, an action should be tried de novo as though there had been no previous trial. (See People v. Simmons (1916), 274 Ill. 528, 113 N.E. 887; Indiana Harbor Belt R.R. Co. v. Green (1919), 289 Ill. 81, 124 N.E. 298; see 28 Ill. L. & Prac. New Trial §Ill (1957).) This is especially important in a criminal prosecution as information about a defendant’s conviction in a former trial is “perhaps even more damaging” than information about prior criminal acts. (See United States v. Williams (5th Cir. 1978), 568 F.2d 464, 470.) An obvious example of judicial concern in this area is the inadmissibility of a prior plea of guilty upon a substituted plea of not guilty (Kercheval v. United States (1927), 274 U.S. 220, 71 L. Ed. 1009, 47 S. Ct. 582) reflected in Supreme Court Rule 402(f) (Ill. Rev. Stat. 1977, ch. 110A, par. 402(f)).
To this judicial concern, we must also add the standard of the more usual consideration of the prejudicial effect of a reference to prior criminality. Subject to the previously mentioned exceptions, it is generally the rule that evidence of a distinct substantive offense, or evidence which tends to mislead or prejudice the jury, cannot be admitted in a prosecution for another offense. (People v. Ciucci (1956), 8 Ill. 2d 619, 137 N.E.2d 40, affd sub nom. Ciucci v. Illinois (1958), 356 U.S. 571, 2 L. Ed. 2d 983, 78 S. Ct. 839. When evidence of prior criminality is not direct, but merely inferential, the determinative issue is the probative and prejudicial effect of the nexus of the admitted and prior criminality. People v. Price (1979), 76 Ill. App. 3d 613, 394 N.E.2d 1256; People v. Coleman (1974), 17 Ill. App. 3d 421, 308 N.E.2d 364.
In the case at bar, the two references to prior criminality were Singleton’s statements that he had seen the defendant at Stateville and that he had talked with defendant “about the Appellate Court, something about his appeal or something.” The former statement clearly inferred prior criminality while the latter statement may have so inferred. The witness did not elaborate on the circumstances or time period during which the conversation took place nor did the People attempt to elicit such information from Singleton. No specific crime was ever mentioned or alluded to, and the People forwarded no further comment, question, or argument concerning the colloquy. Thus, while an inference of prior criminality was presented by the testimony, the inference was nebulous and remained ambiguous throughout the trial.
*461We also note that the jury was clearly aware of defendant’s prior criminality before the occurrence of the colloquy in question. An important circumstance here present is the fact the homicide took place while defendant was incarcerated. Of less importance is the fact the jury was equally aware the matter had taken almost seven years to come to trial, a circumstance which may well have inferred previous legal proceedings.
Considering the nebulous and ambiguous nature of the testimony at bar, coupled with the jury’s prior awareness of defendant’s incarceration and the lapse of time before trial, the probative effect of Singleton’s testimony was greater than its prejudicial effect upon the jury. We note, however, that a new trial is here involved and, given the judicial concern regarding information about prior proceedings in this situation, we shall consider the evidence adduced at trial in light of the possible prejudicial effect of Singleton’s testimony.
At trial, the People called two individuals who were inmates at the Industrial School at the time of the incident and were eyewitnesses to the stabbing. The first inmate stated that he saw defendant stab Jefferson in the back with a black object wrapped in black tape. The second inmate stated he had seen defendant with a sharpened comb without its teeth and subsequently saw defendant stab the deceased in the neck or chest with a long, dark object. Defendant’s commission of the homicide was further corroborated by the testimony of two employees of the school. The supervisor of the building where the incident took place and a guard each stated that Jefferson had named defendant as one of his assailants before he died.
The defense called three inmates, including Singleton, and the defendant. Singleton and one inmate stated that defendant did not stab Jefferson but added on cross-examination that they did not see anyone stab the deceased guard. The third inmate testified that he did not see defendant outside of his cell. Defendant stated he remained in front of his cell and Singleton’s cell and made no move toward Jefferson at any time.
For an error to be harmless, the court must be convinced beyond a reasonable doubt that it did not contribute to the verdict. (Chapman v. California (1967), 386 U.S. 18, 17 L. Ed. 2d 705, 87 S. Ct. 824.) Considering the testimony of two eyewitnesses and two additional witnesses who heard the victim name defendant as an attacker, coupled with the fact the defense witnesses did not actually observe the stabbings, we find the evidence introduced at trial establishes the defendant’s guilt beyond a reasonable doubt and the error was harmless under the Chapman test. This being the case, any error resulting from Singleton’s testimony is harmless and does not require reversal.
Defendant’s second contention is that his 20- to 40-year sentence is excessive as the evidence at his resentencing hearing indicated his *462rehabilitation since he originally received a similar sentence. Under Supreme Court Rule 615(b) (4) (Ill. Rev. Stat. 1977, ch. 110A, par. 615(b)(4)), the reviewing courts may reduce the sentence imposed by the trial court. As the sentencing hearing herein took place January 27,1978, we need not consider the standard of review established by section 5 — 5—4.1 of the Unified Code of Corrections (Ill. Rev. Stat., 1978 Supp., ch. 38, par. 1005 — 5—4.1) or its appropriateness. See People v. Godinez (1980), 81 Ill. App. 3d 547, 401 N.E.2d 658 (dissenting opinion).
The imposition of a sentence is a matter of judicial discretion and absent an abuse of this discretion, the sentence may not be altered upon review. (E.g., People v. Perruquet (1977), 68 Ill. 2d 149, 368 N.E.2d 882.) The trial court is in a better position to determine the punishment to be imposed and consider the particular circumstances of each individual case, fashioning a sentence which protects societal interests while allowing for offender rehabilitation. E.g., People v. Perruquet.
The trial court considered information contained in the presentence investigation report, psychiatric reports, and defendant’s written statement. The presentence report indicated juvenile adjudications of armed robbery and involuntary manslaughter and that defendant had been in penal institutions since the age of 14. The psychiatric reports evidence intellectual, psychological, and social development during defendant’s incarceration, and there is no question that defendant made great strides in rehabilitating himself. Defendant’s written statement asked the court to consider this rehabilitation in light of his desire to counsel young people.
Contraposed against defendant’s rehabilitative success is the societal interest reflected in the death penalty being applicable to crimes of this nature under section 9 — 1(b)(2) of the Criminal Code of 1961. (Ill. Rev. Stat. 1977, ch. 38, par. 9 — 1(b)(2).) The murder of a correctional officer is one of the most heinous crimes which society has seen fit to proscribe. The spirit and purpose of the law are upheld when a sentence reflects the seriousness of the offense and the objective of restoring the offender to useful citizenship. (Ill. Const., art. I, §11.) Given the seriousness of the offense at bar, we cannot say the trial court abused its discretion in imposing a sentence of 20 to 40 years.
Accordingly, the judgment of the circuit court of La Salle County is affirmed.
Affirmed.
ALLOY, P. J., and STENGEL, J., concur.