OPINION

DONALD R. ROSS, Justice.

Francis M. Hardin, Jr. was convicted of sexual assault of a child, a felony of the second degree. The victim was his stepdaughter, Kristy Brown (a pseudonym).

After filing an election for the jury to assess punishment and an application for community supervision, Hardin entered a plea of guilty. The trial court duly admonished him and accepted his plea. He again pled guilty before the jury, and at the conclusion of this unitary trial, the court instructed the jury to find Hardin guilty and to assess his punishment within the range prescribed by law. The jury found Hardin guilty and assessed the maximum confinement of twenty years, making him ineligible for community supervision.

Hardin now appeals, alleging that the trial court erred by:

(1) allowing an expert witness to testify without sufficient notice to the defense, in violation of his right to due process;

(2) allowing an expert witness to testify without sufficient notice to the defense, depriving him of effective assistance of counsel;

(3) admitting inadmissible testimony;

(4) admitting testimony of an unqualified expert; and

(5) failing to grant his motion for mistrial after the State asked a witness an improper question.

Much of the controversy in this case centers around the State's use of an expert witness in rebuttal to Hardin's evidence that he would not be a future threat to society if given community supervision. The expert witness, San Thomason, a Bowie County Community Supervision officer who supervises sex offenders released on probation, testified about the characteristics of sex offenders, the most common requirements of community supervision for sex offenders, the likelihood that sex offenders will reoffend, and that in her expert opinion, Hardin fit the profile of a child molester.

In his first two points of error, Hardin contends that the trial court erred by allowing Thomason to testify without sufficient notice to the defense. He contends that such error violates his right to due process and to effective assistance of counsel.

The record reveals that on November 11, 1998, Hardin filed a discovery motion requesting, in part, that the trial court order the State to produce, "A list of names and addresses of the witnesses to be called by the prosecution in this instant case, and whether or not they are to be called as direct or rebuttal witnesses." The clerk's record does not reveal whether the trial court ruled on the defendant's motion, but at trial both the State's attorney and Hardin's attorney discussed a hearing, apparently on January 14, at which the judge ordered the State to provide a witness list.

On Friday, January 15, the State provided a witness list to Hardin's attorney. This list did not include Thomason's name. On the same day, but apparently after the State sent its witness list, Hardin filed his application for community supervision. On Tuesday morning, January 19, the first working day of that week, Hardin entered his plea of guilty and was duly admonished by the court. A jury was selected on the afternoon of the same date. Trial was then recessed until Thursday, January 21.

On Wednesday, January 20, the State notified Hardin of its intent to call Thomason as a witness. During trial the next day, the State called Thomason in rebuttal to Hardin's witnesses, who testified that he would not be a future danger to society. Hardin objected on the ground that he did not have the opportunity to voir dire the jury about Thomason.

Hardin contends that the State knew on January 15, the date he filed his application for community supervision, that he would put on evidence about his suitability for community supervision. He further contends that the State knew at least by January 19 that it would call Thomason as a witness and should have informed Hardin on that date before the jury was selected. He also argues that he could not have anticipated the State calling Thomason because Thomason had never met Hardin or interviewed him.

The State contends that it notified Hardin's attorney "as soon as it had determined [Thomason] would be a potential witness, after [Hardin] changed the course of the trial at the last minute." It also contends that Hardin should have expected the State to call a rebuttal witness to counter his claim that he is suitable for community supervision.

The State should disclose witnesses if they will be used at any stage in the trial. Hightower v. State, 629 S.W.2d 920, 925 (Tex. Crim App. [Panel Op.] 1981); Young v. State, 547 S.W.2d 23, 27 (Tex.Crim.App. 1977). The State also has a continuing burden of disclosure once the trial court grants a discovery motion. Crane v. State, 786 S.W.2d 338, 348 (Tex.Crim.App. 1990).

If the trial court allows a witness who is not on the State's list to testify, we review that decision for abuse of discretion. Martinez v. State, 867 S.W.2d 30, 39 (Tex.Crim.App. 1993); Lafayette v. State, 835 S.W.2d 131, 132-33 (Tex. App.-Texarkana 1992, no pet.). Our review encompasses two factors: (1) whether the State's actions constituted bad faith, and (2) whether Hardin could have reasonably anticipated that Thomason would testify. Nobles v. State, 843 S.W.2d 503, 514-15 (Tex.Crim.App. 1992); Lafayette, 835 S.W.2d at 132-33. Any error in allowing the witness to testify over a claim of surprise is made harmless by the defendant's failure to object or move for a continuance. Barnes v. State, 876 S.W.2d 316, 328 (Tex.Crim.App. 1994).

In determining whether the State acted in bad faith, reviewing courts have examined at least three areas of inquiry:

(1) whether the defense shows that the State intended to deceive; see Nobles, 843 S.W.2d at 515; Richardson v. State, 744 S.W.2d 65, 78 (Tex.Crim.App. 1987), vacated remanded on other grounds, 492 U.S. 914, 109 S.Ct. 3235, 106 L.Ed.2d 583 (1989);

(2) whether the State's notice left the defense adequate time to prepare; see Hernandez v. State, 819 S.W.2d 806, 816 (Tex.Crim.App. 1991); Nobles, 843 S.W.2d at 515; and

(3) whether the State freely provided the defense with information (e.g., by maintaining an open files policy, by providing updated witness lists, or by promptly notifying the defense of new witnesses). See Richardson, 744 S.W.2d at 78; Pinkerton v. State, 660 S.W.2d 58, 64 (Tex.Crim. App. 1983).

Similarly, in determining whether the defense could have anticipated the State's witness, reviewing courts also have examined three areas of inquiry:

(1) the degree of surprise to the defendant; see Martinez, 867 S.W.2d at 39; Nobles, 843 S.W.2d at 515; Hightower, 629 S.W.2d at 925;

(2) the degree of disadvantage inherent in that surprise (e.g., the defendant was aware of what the witness would say, or the witness testified about cumulative or uncontested issues); see Hernandez, 819 S.W.2d at 816; Goodley v. State, 457 S.W.2d 294, 296 (Tex.Crim.App. 1970); and

(3) the degree to which the trial court was able to remedy that surprise (e.g., by granting the defense a recess, postponement, or continuance, or by ordering the State to provide the witness' criminal history). See Hernandez, 819 S.W.2d at 816; Stoker v. State, 788 S.W.2d 1, 15 (Tex.Crim. App. 1989); Richardson, 744 S.W.2d at 77; Bridge v. State, 726 S.W.2d 558, 566-67 (Tex.Crim.App. 1986); Pinkerton, 660 S.W.2d at 64.

From the record, it does not appear that the State acted in bad faith. The record does not clearly establish that the State knew that it intended to call Thomason as a rebuttal witness and willfully refused to disclose it to Hardin. The State informed Hardin that it intended to call Thomason to testify two working days after Hardin filed his application for community supervision, and one day before commencement of the trial following jury selection. Hardin presented no evidence to refute the State's assertion that this notice was "as soon as it had determined [Thomason] would be a potential witness. . . ." At the time Hardin received this notice, he still had one day to prepare for his cross-examination and could have requested a recess or continuance if he needed more time.

Deciding whether Hardin could have anticipated that the State would call Thomason is more problematic. Hardin should have been able to anticipate that the State would attempt to rebut any testimony concerning his suitability for community supervision. Again, he also did not request more time to prepare once he learned of the State's intention to call Thomason.

In contrast to almost every other Texas case on this issue, however, Hardin does not claim that he was disadvantaged by having insufficient time to prepare for cross-examination; rather, he contends that he was unable to voir dire the jury panel about Thomason. In the only two cases to address this issue directly, the reviewing court affirmed the trial court's decision to allow the witness to testify. Hernandez, 819 S.W.2d at 816; Salinas v. State, 625 S.W.2d 397, 401-02 (Tex. App.-San Antonio 1981, no pet.). In Hernandez, the court's analysis centered around the absence of evidence showing that the State acted in bad faith. Hernandez, 819 S.W.2d at 816. In Salinas, the court cited the lack of evidence showing the defendant was harmed, whether by a motion for new trial, bill of exception, or otherwise. Salinas, 625 S.W.2d at 401-02. The court also looked at the cumulative nature of the witness' testimony. Id.

It is important to note that the inability to voir dire the jury about a surprise witness is inherent in most, if not all, such cases. However, "[t]o require the State to anticipate any possible defense of an accused and to furnish names of all possible witnesses and have the court refuse to permit them to testify if their names were not listed would be to require an impractical and undue burden." Hoagland v. State, 494 S.W.2d 186, 189 (Tex.Crim.App. 1973).

In this case, Hardin does not show how the trial court's decision disadvantaged him. Instead, he alleges general disadvantage associated with being unable to determine whether and to what extent the venire members would be influenced by an expert's testimony and whether any of the venire members knew Thomason. However, once Hardin filed his application for community supervision, it was reasonable to anticipate that the State would offer testimony from someone — if not this particular witness — to rebut Hardin's evidence concerning his suitability for community supervision, if he opened that door. Further, at no time did Hardin's attorney attempt to elicit evidence of harm at trial (e.g., by asking Thomason outside the presence of the jury if she had any connection with any of the jurors); Absent a clear showing of actual harm or bad faith, we hold that the trial court did not abuse its discretion. Hardin's first two points of error are overruled.

In his third point of error, Hardin contends that the trial court erred "by allowing . . . Thomason to testify as an expert regarding [Hardin's] suitability for probation." As we understand his contention under this point of error, he is not contending that the trial court should have disallowed Thomason's testimony as an expert. See Hardin's fifth point of error, infra. Rather, he is contending that Thomason's testimony was inadmissable for the purpose for which the State offered it, specifically, to rebut his own evidence that he was suitable for community supervision.

At the punishment phase, Hardin called Christy Williams (his ex-wife) and Steve Hardin (his brother) as witnesses on his behalf. Both Williams and Steve Hardin testified that they believe Hardin does not pose a threat to the community. Hardin then took the stand in his own defense. When his attorney asked him what message he would like to give the jury, he responded, "I would ask, if it was at all possible, for the jury to grant me . . . community supervision. You would not be sorry. This would never happen again."

Hardin cites Ortiz v. State, 834 S.W.2d 343, 346 (Tex.Crim. App. 1992), contending that evidence tending to prove that an applicant for community supervision is or is not suitable for community supervision is objectionable. Hardin made a timely objection to Thomason's testimony. The State cites the same case to support its contention that a party may open the door to evidence of a defendant's suitability for community supervision.

We review the trial court's decision to admit or exclude evidence under an abuse of discretion standard. Green v. State, 934 S.W.2d 92, 101-02 (Tex.Crim.App. 1996); Taylor v. State, 995 S.W.2d 279, 283 (Tex. App.-Texarkana 1999, pet. granted). We look to see whether the court acted without reference to any guiding rules or principles. Lyles v. State, 850 S.W.2d 497, 502 (Tex.Crim.App. 1993). The mere fact that a trial court may decide a matter within its discretionary authority differently than a reviewing court does not demonstrate such an abuse. See Montgomery v. State, 810 S.W.2d 372, 391-92 (Tex.Crim.App. 1990) (opinion on reh'g). Therefore, we will not reverse a trial court whose ruling was within the "zone of reasonable disagreement." Green, 934 S.W.2d at 102; Taylor, 995 S.W.2d at 283.

A defendant's suitability for probation or his ability to abide by the terms and conditions of probation are not issues in the punishment phase of trial. Ortiz, 834 S.W.2d at 346; Griffin v. State, 787 S.W.2d 63, 67 (Tex.Crim.App. 1990). Any evidence that a defendant propounds to show his suitability for probation is objectionable. Ortiz, 834 S.W.2d at 346; Griffin, 787 S.W.2d at 67. However, if the defendant opens the door by offering such evidence, the State may choose to forego an objection and offer rebuttal evidence instead. Ortiz, 834 S.W.2d at 346; Griffin, 787 S.W.2d at 67.

In this case, by offering testimony that he would not be a future danger to society, Hardin opened the door to the kind of evidence that Thomason would provide, i.e., the likelihood that a sex offender will reoffend and therefore constitute a future danger to society. By allowing this rebuttal evidence, the trial court acted within the zone of reasonable disagreement. We overrule Hardin's third point of error.

In a related point of error, Hardin contends that the trial court erred by allowing Thomason to testify as an expert. We review the trial court's decision to admit expert testimony for abuse of discretion. Alvarado v. State, 912 S.W.2d 199, 216 (Tex.Crim. App. 1995); Chisum v. State, 988 S.W.2d 244, 250 (Tex. App.-Texarkana 1998, pet. ref'd).

TEX. R. EVID. 702 governs the admission of expert testimony. The test for admissibility of expert testimony is (1) the expert must be qualified by knowledge, skill, expertise, training, or education; (2) the subject matter of the testimony must be an appropriate one for expert testimony; and (3) the expert's testimony must assist the trier of fact to understand the evidence or decide a fact in issue. Alvarado, 912 S.W.2d at 215-16; Bui v. State, 964 S.W.2d 335, 344 (Tex. App.-Texarkana 1998, pet. ref'd). Even if the expert is qualified and the subject matter is appropriate for expert testimony, however, the trial court must also be satisfied that the evidence is both relevant and reliable. Hartman v. State, 946 S.W.2d 60, 62 (Tex.Crim. Apr. 1997) (citing Kelly v. State, 824 S.W.2d 568, 572 (Tex.Crim. App. 1992)); see also 2A STEVEN GOODE ET AL., TEXAS PRACTICE: COURTROOM HANDBOOK ON TEXAS EVIDENCE Rule 702, cmt. 6 (1999).

Thomason testified outside the presence of the jury to the following: that she has been employed with Bowie County Community Supervision for twelve years; that before working for Bowie County, she worked for the Texas Department of Human Services for eight years; that her job with Bowie County has been to supervise sex offenders who are on probation; that her job with the Department of Human Services was to investigate cases of child abuse and neglect, focusing specifically on cases involving sex offenders; that she has a bachelor of science degree in psychology and sociology and twenty-one hours toward her master's degree as a psychological associate; that she receives at least forty hours per year of training in the subject of sexual deviancy; that she has received this training for the past twenty years; that she recently received training about sex offenders at the National Institution of Corrections, Department of Justice Academy and has been chosen to receive training from the United States Marshals about sex offenders and the internet; that she has testified as an expert over 400 times in her twenty-year career; that she has conducted studies of recidivism rates of sex offenders in Bowie County, but has not conducted any national studies; that her training has covered national recidivism statistics provided by the Department of Justice and other agencies; and that experts in her field use those statistics to make assessments and determinations.

Hardin contends that Thomason's testimony does not meet the reliability test outlined in Kelly, 824 S.W.2d at 572, and its progeny. In Kelly, the Texas Court of Criminal Appeals held that scientific evidence is admissible if (1) the underlying scientific theory is valid, (2) the technique applying the theory is valid, and (3) the technique was properly applied. Id. at 573.

Hardin contends the State did not establish that Thomason had any theories underlying her testimony, or that those theories are accepted as valid by the scientific community. He points out that Thomason did not testify that any literature existed to support her underlying theories or that her theories had been empirically tested. He also contends that she did not testify about the nature and extent of her independent research about sex offenders and their rate of recidivism. In particular, Hardin complains that Thomason was not qualified to testify about whether Hardin fits the profile of a child molester and about recidivism rates.

Hardin's reliance on Kelly is incomplete, however. In Nenno v. State, 970 S.W.2d 549, 561 (Tex.Crim.App. 1998), the Texas Court of Criminal Appeals stated "an appropriately tailored translation of the Kelly test" for nonscientific expert testimony. When the expert is from a nonscientific discipline (i.e., one involving technical or other specialized knowledge, experience, and training as opposed to the scientific method), the test for reliability is (1) whether the field of expertise is a legitimate one, (2) whether the subject matter of the expert's testimony is within the scope of that field, and (3) whether the expert's testimony properly relies on and/or utilizes the principles involved in the field. Id.

In Nenno, the court applied this test to evaluate the reliability of expert testimony from a special agent for the FBI who specialized in studying the sexual victimization of children. Id. at 562. The State called the agent to discuss the issue of future dangerousness in a death penalty case. Id. The court found that through interviews, case studies, and statistical research, a person could acquire superior knowledge concerning the behavior of such offenders. Id. In particular, the court found that the issue of future dangerousness is one that often surfaces in the course of research on sex offenders. Id. The court also observed that the lack of peer review for such research would go to the weight of the evidence and not its admissibility. Id.

In the present case, Thomason was qualified to testify as an expert on the issue of whether a convicted sex offender poses a continuing threat to society. It is true that Thomason testified to conducting only limited research in the area of recidivism rates. However, her twenty years' experience investigating and supervising sex offenders, and her extensive and continuous training qualified her to testify as an expert under the "tailored" standards of Nenno. Hardin's fifth point of error is overruled.

Hardin also contends that the trial court erred by failing to grant his motion for mistrial after it sustained his objection to one of the State's questions. At trial, the State called Steven Tyler, Hardin's coworker, and during direct examination, the following exchange occurred:

Q. Mr. Tyler, as the head of janitorial services while the Defendant was working out there [an elementary school], did you ever get a report from any of your janitors regarding the Defendant and a fifth grader?

A. Yes.

[DEFENSE COUNSEL]: Judge, I'm going to object. It'd be hearsay for a report that he's seen. They're not trying to offer any other exception. It's, did you get a report, which is clearly hearsay.

THE COURT: Approach the bench.

[PROSECUTOR]: I'll withdraw the question.

[DEFENSE COUNSEL]: Judge, I would request that you instruct the jury just to disregard the question.

THE COURT: The jury is instructed to disregard the question. There is no answer to it. Completely disregard it.

[DEFENSE COUNSEL]: Your Honor, may I approach?

THE COURT: Yes, sir.

(Bench conference out of hearing of jury.)

[DEFENSE COUNSEL]: He has placed the implication of inappropriate conduct regarding a fifth grader, and I don't think your correcting instruction will cure the harm, and move for a mistrial.

THE COURT: The objection is overruled — no. The objection is not overruled. The motion is denied.

Under the circumstances of this case, the inference that could be drawn from the question asked clearly made it an improper question. This was apparently recognized by the prosecutor, who promptly withdrew it. We review a trial court's denial of a mistrial under an abuse of discretion standard. Ladd v. State, 3 S.W.3d 547, 567 (Tex.Crim.App. 1999). Mistrial is an extreme remedy, and reviewing courts rarely reverse a trial court for refusing to grant a mistrial after an improper question. See Bauder v. State, 921 S.W.2d 696, 698 (Tex.Crim.App. 1996); see also Swallow v. State, 829 S.W.2d 223, 226-27 (Tex.Crim.App. 1992). Most often, error from an improper question can be cured or rendered harmless by its withdrawal or an instruction to disregard. Hernandez v. State, 805 S.W.2d 409, 413-14 (Tex.Crim. App. 1990). The exception occurs in extreme cases where the question is clearly calculated to inflame the minds of the jury and is of such a character as to suggest the impossibility of withdrawing the impression produced. Id. at 414.

Whether a curative instruction is effective depends on the facts and circumstances of each case. Swallow, 829 S.W.2d at 227. The question is essentially one of balancing the magnitude of the probable effect of the question on the jury against the probable efficacy of an instruction to disregard. The greater the magnitude of effect, the less likely it is that an instruction would cure the harm.

The State cites Bauder in contending that the curative instruction was sufficient to remedy any prejudice from the improper question. In Bauder, the Texas Court of Criminal Appeals observed:

Even when a prosecutor intentionally elicits testimony or produces other evidence before the jury which is excludable at the defendant's option, our law prefers that the trial continue. Because tactical decisions to offer prejudicial evidence are a normal and, in most respects, acceptable part of the adversary process, it would be counterproductive to terminate the trial every time an objection is sustained. Consequently, it is considered a sufficient response to most well-founded objections that the material be withdrawn from jury consideration, if necessary, and that jurors be admonished not to consider it during their deliberations.

Bauder, 921 S.W.2d at 698 (citations omitted).

Bauder should not be read, however, as sanctioning the conduct of zealous advocates who skate close to the ragged edge of improperly prejudicing the jury. Rather, the court's comments must be understood as expressing a strong preference for completing the trial, as long as the trial remains fair. As the court commented elsewhere in Bauder: "[o]nly when it is apparent that an objectionable event at trial is so emotionally inflammatory that curative instructions are not likely to prevent the jury being unfairly prejudiced against the defendant may a motion for mistrial be granted." Id.

We have reviewed numerous Texas Court of Criminal Appeals cases regarding curative instructions and determined that the court has found several nonexhaustive, nonexclusive factors important in assessing whether an instruction to disregard will cure the harm:

(1) The weight of other evidence supporting the verdict. Waldo v. State, 746 S.W.2d 750, 754-57 (Tex.Crim.App. 1988); see also Hinojosa v. State, 4 S.W.3d 240, 253 (Tex.Crim. App. 1999); Whitaker v. State, 977 S.W.2d 595, 600 (Tex.Crim. App. 1998).

(2) The nature and form of the question (e.g., whether the question was stated as a fact designed to elicit specific evidence of a specific extraneous offense or bad act). See Ladd, 3 S.W.3d at 566-67; Adanandus v. State, 866 S.W.2d 210, 227 (Tex.Crim.App. 1993); Swallow, 829 S.W.2d at 227; Hernandez, 805 S.W.2d at 413-14; Legg v. State, 594 S.W.2d 429, 433 (Tex.Crim.App. [Panel Op.] 1980).

(3) The State's purpose and persistence in asking the question. See Burks v. State, 876 S.W.2d 877, 902 (Tex.Crim.App. 1994); Boyd v. State, 811 S.W.2d 105, 122 (Tex.Crim.App. 1991); Waldo, 746 S.W.2d at 754-57; Crawford v. State, 603 S.W.2d 874, 876 (Tex.Crim.App. [Panel Op.] 1980); Carrillo v. State, 591 S.W.2d 876, 892-93 (Tex.Crim.App. [Panel Op.] 1979); Cavender v. State, 547 S.W.2d 601, 603 (Tex.Crim.App. 1977).

(4) Whether other evidence concerning the same subject has been admitted. See Huffman v. State, 746 S.W.2d 212, 218-19 (Tex.Crim.App. 1988); Yarbrough v. State, 617 S.W.2d 221, 228 (Tex.Crim.App. [Panel Op.] 1981); see also Ransom v. State, 789 S.W.2d 572, 585 (Tex.Crim.App. 1989).

(5) The particular instruction given. Waldo, 746 S.W.2d at 754-57; see also Ladd, 3 S.W.3d at 567; Swallow, 829 S.W.2d at 227; Davis v. State, 645 S.W.2d 817, 819 (Tex.Crim. App. 1983).

(6) The harm to the accused as measured by the severity of the sentence. Waldo, 746 S.W.2d at 754-57; see also Gonzales v. State, 685 S.W.2d 47, 49-50 (Tex.Crim.App. 1985).

We proceed to evaluate the trial court's instruction to disregard using these factors.

Weight of the Evidence/Severity of Punishment

In assessing the weight of the evidence, we note first that Hardin pled guilty to sexual assault of a child, an offense reasonably expected to elicit strong feelings in most jurors. The State's chief witness was the victim, Kristy Brown. She testified to a five year progression of inappropriate sexual contact, beginning when she was twelve and ending when she and her mother left Hardin when she was about seventeen. This sexual contact began with inappropriate touching, which Kristy testified occurred daily when she was twelve and thirteen. It progressed to Hardin requiring her to stimulate him manually, which she testified occurred weekly and began when she was thirteen years old.

Kristy also testified that Hardin required her to give him oral sex, simulated sex with her, once told her to cut off her pubic hairs and give them to him, and requested to have sex with her. She testified that if she refused his overtures, he would treat her and her mother harshly until she agreed. Kristy's mother, Hardin's ex-wife, corroborated key elements of her daughter's story, including Hardin's sexual patterns and his harsh behavior.

The State called Tyler to recount conversations he had with Hardin. Before the State asked the improper question at issue, Tyler testified that Hardin told him that when he was nineteen, he had sex with his first wife's sister, who was fourteen years old (Hardin was fifty-four years old at the time of trial).

In his direct examination, Hardin essentially disputed only the duration and frequency of his sexual contact with Kristy, but did not dispute that it occurred. He contended that the things he told Tyler were exaggerations made in the context of a working relationship where they shared numerous "tall stories." He told the jury that this was the first time he had ever been in trouble and had never even had a traffic ticket.

In her rebuttal testimony, Thomason discussed at length the characteristics of child molesters, who prey upon post-pubescent children. She further testified that child molesters typically molest children in their homes. They begin, she stated, with seemingly innocent touching and progress from there, using emotional manipulation to enforce their demands and maintain control. She then stated that in her expert opinion after reviewing the case file, Hardin fit the profile of a child molester. She later testified that sex offenders are highly likely to reoffend.

Hardin received twenty years' imprisonment, the maximum number of years allowable. This alone is not necessarily dispositive of harm from an improper question, however. In Waldo, 746 S.W.2d at 757, where the defendant received the maximum sentence, the court looked at the defendant's sentence in terms of the offense and the other evidence in deciding that he was not harmed by the improper question. In this case, there was ample evidence from which the jury could have concluded Hardin did not deserve community supervision and did deserve the maximum years' imprisonment.

Nature and Form of the Question

Some courts have stated that "whenever the question is so stated that it amounts to an assertion of a fact . . . and it implies the commission of another offense, it may be said that its harmfulness cannot be cured by the answer and seldom by any instruction which the court is able to give the jury." Swallow, 829 S.W.2d at 227 (quoting McNaulty v. State, 138 Tex.Crim. 317, 135 S.W.2d 987, 988-89 (1939) (opinion on reh'g)). However, even when the improper question directly confronts the jury with a factual assertion that the defendant has committed a specific inadmissible extraneous offense, courts have held that the instruction to disregard cured the harm. Id.; see also Ladd, 3 S.W.3d at 571.

Hardin contends that the question was calculated to correspond with the State's strategy of portraying him as a deviate. As a result, he contends the question left the jury with the lingering and unerasable impression that he had harmed more than one child in the past and, therefore, would harm another child in the future.

At oral argument, Hardin also emphasized the fact that Tyler was able to answer the question before his attorney objected. It is true that at times courts phrase the rule regarding the efficacy of curative instructions in terms of whether the witness answered the question. See, e.g., Nenno, 970 S.W.2d at 563 ("Generally, when a witness has not had the opportunity to answer an improper question, an instruction to disregard will cure any harm to the defendant."). Other times, however, the fact that the witness answered the improper question is of little importance to the court. Ladd, 3 S.W.3d at 567.

We think that whether the witness answers the question is important primarily for the information the answer reveals. Our understanding corresponds with the consistency between the rules in those cases where the State asks an improper question, and in those cases where the witness' nonresponsive answer to a proper question brings in inadmissible evidence. The rule in the latter cases is the same as in the former (i.e., a curative instruction is presumed effective except in extreme cases when the question is clearly calculated to inflame the minds of the jury and is of such a character as to suggest the impossibility of withdrawing the impression produced). See, e.g., Hughes v. State, 878 S.W.2d 142, 154 (Tex.Crim.App. 1992) (opinion on reh'g).

In this case, for example, any prejudice was derived from the nature of the question. The State's question was prejudicial insofar as it implied that Hardin had improper contact with another child. Tyler's answer did not add much in the way of prejudice because, even if he had not answered, the jury could have derived the prejudicial inference. That is not to say that the question was so prejudicial that the trial court's instruction could not have cured it. As the State rightly observes, its question did not indicate the nature of the report and did not directly reference any crime or bad acts.

The State's Purpose and Persistence/other Evidence About the Same Subject

From the record, it appears that the question to Tyler was designed to leave the jury with the impression that Hardin had inappropriate contact with another child. Moreover, we do not perceive a permissible purpose for asking the question, and the State has offered none. Several cases have cited the State's bad faith as an important factor in gauging the magnitude of harm caused by an improper question. See, e.g., Cooks v. State, 844 S.W.2d 697, 735 (Tex.Crim.App. 1992); Walker v. State, 610 S.W.2d 481, 483 (Tex.Crim.App. [Panel Op.] 1980); Cavender, 547 S.W.2d at 603.

However, we note that the State promptly withdrew the question and did not persist in raising the subject. No other evidence about this matter reached the jury during trial. Hardin contends that the State referred to this exchange in its closing argument when the prosecutor stated, "It's true that he's had inappropriate sexual conduct with children before. And folks, I think it's true that it'll happen again." In context, however, the State's argument clearly refers to elements of Thomason's testimony and other aspects of Tyler's testimony, including Tyler's testimony that Hardin told him about having sex with his ex-wife's fourteen-year-old sister.

The Trial Court's Instruction

Immediately after the State withdrew its question, Hardin's attorney requested the court to instruct the jury to disregard the question. The court promptly instructed the jury to "disregard the question" and to "[c]ompletely disregard it." The court also instructed the jury that there was "no answer" to the question, which we interpret to mean that they were to disregard the answer as well.

Application

Weighing all of these factors, we cannot say that the trial court abused its discretion in refusing to order a mistrial. There was ample evidence before the jury from which it could have concluded that Hardin did not deserve community supervision, but did deserve assessment of the maximum number of years' confinement. In the context of the record, this was one isolated question that, though improper and prejudicial, did not refer directly to an extraneous offense. The State did not pursue the issue beyond the question. The court gave a prompt and thorough instruction to disregard both the question and the answer. We are confident the court's instruction cured any resultant harm. See Ladd, 3 S.W.3d at 571. We overrule this point of error.

The judgment of the trial court is affirmed.