dissenting.
There are three portions of the majority opinion with which I disagree. Two of them relate to the majority opinion’s analysis of a comment made by the prosecutor; first, whether there was error, and second, if error, was it harmless. The third portion of the opinion with which I disagree is its analysis of the admission of testimony as lay opinion rather than expert opinion. I will address the issue regarding the testimony first and then address the analysis of the comment by the prosecutor.
ExpeRt vs. Lay Opinion Testimony
Roberson’s second issue is that the testimony of a police officer about the normal conduct of sexual assault victims was improperly admitted. Roberson argues that to be admissible the testimony must meet the requirements of Rule 702 regarding the admission of expert testimony. The State argues that it is not expert testimony and should be analyzed as lay opinion testimony under Rule 701. A majority of the court has followed the State’s argument and analyzed it as lay opinion testimony.
*46I would analyze it as expert opinion testimony. Actually, I would assume without deciding that the admission was error, proceed directly to a harm analysis, and hold that the admission of this testimony from Lanning, if error, was harmless. Tex.R.App. P. 42.2(b).
Effect of InstRüction to Diseegaed
In its analysis of Roberson’s first issue, a majority of the court determines that the prejudicial effect of an improper comment by the prosecutor was not cured by an instruction to disregard. I disagree.
First, the comment was not so blatant that an instruction could not cure the prejudicial effect, if any. See Waldo v. State, 746 S.W.2d 750, 753 (Tex.Crim.App.1988). The Court of Criminal Appeals has noted that they tend to find instructions to disregard improper argument to have the desired effect of removing the improper argument from the consideration of the jury in all but the most blatant cases of commenting on the defendant’s failure to testify Id. The State was attempting to respond to Roberson’s argument that he was a good candidate for probation with the concept of admitting a problem is a necessary step to changing bad habits. The prosecutor impliedly argued that because Roberson pled guilty, he was not willing to change. The argument was improper, but not a blatant comment on Roberson not testifying.
Second, the trial court’s instruction was not faulty. Yes, the prosecutor was not instructed to cease her line of argument, but there is no requirement that the trial court do so. An instruction specifically informing the jury to disregard the last comment made has been held to be adequate. See Waldo, 746 S.W.2d at 755-756.
Third, the majority glossed over the weight of the incriminating evidence. Even if we must only consider the weight of the punishment evidence, that consideration necessarily includes the evidence of guilt. See Tex.Code Ceim. PROC. Ann. art. 37.07, § 3(a)(1) (Vernon Supp.2002). Roberson and the victim at times shared various residences with other people. They were not in a relationship with each other. Roberson sexually assaulted his victim six times between the end of 1997 and the end of 1998. When these offenses began, the victim was only 15 years old. All six times, Roberson assaulted the victim while she was asleep. After she awoke, he would stop and leave. Twice he pinned her down and held her teddy bear over her face. She struggled with him to get him to leave.
At punishment, the State introduced evidence that Roberson sexually assaulted another woman while asleep in the victim’s bed. He also sexually assaulted this woman’s six-year old daughter. At the time of this trial, an indictment was pending against Roberson for the assault of the daughter. Roberson also physically abused his girlfriend’s two-year old daughter. The girlfriend initially denied to authorities that Roberson committed the abuse because he threatened to kill her and her daughter. Roberson continually threatened the girlfriend with harm, threatened to “shoot up” her house, and used weapons which scared her. The girlfriend stated that Roberson had a bad reputation for being peaceful and law abiding.
The State also introduced evidence that Roberson frightened, offended, and intimidated a neighbor while she was cleaning up a yard of a vacant house. Roberson yelled obscenities and made obscene gestures to the woman to such an extent that she ran home and called the police. Roberson was apprehended a short time later. Additionally, the police officer who arrested Roberson for the case at hand, *47smelled the odor of burned marijuana when he arrested Roberson and believed Roberson to be under the influence of an illegal substance.
Roberson introduced the testimony of his sister and mother who said they would help him with probation if the jury would give him a chance. Roberson’s sister had just completed probation for “fighting.” She denied ever threatening any of the witness in the case. On rebuttal, the State brought forth two witnesses who were either threatened with death by Roberson’s sister if caught lying about Roberson or who heard her threaten the victim with death if she were lying about the abuse or if Roberson was convicted.
Fourth, 17 years for this sexual assault based on the punishment evidence is not unduly severe. The maximum is 20 years. The jury did not even assess the maximum fine.
In weighing the evidence associated with each of the factors, the instruction to disregard was adequate to cure the prejudicial effect of the improper argument caused by the State to the extent that I am certain the impression produced on the minds of the jury was withdrawn. See Waldo, 746 S.W.2d at 754 (Tex.Crim.App.1988). Compare Chimney v. State, 6 S.W.3d 681, 703-04 (Tex.App.-Waco 1999, no pet.)(trial court’s instruction to disregard cured prejudicial effect of State’s comment on defendant’s failure to testify during argument). I would hold that the trial court did not err in denying Roberson’s request for a mistrial, and no harm analysis would be necessary.
Harm Analysis
After determining that the prejudicial effect was not cured by the instruction and therefore holding the trial court erred in failing to grant a motion for new trial, the majority conducts a harm analysis. The harm analysis is defective because it fails to fully consider “... evidence of the defendant’s guilt ...” Motilla v. State, 78 S.W.3d 352, 357 (Tex.Crim.App.2002). The Court of Criminal appeals has recently reaffirmed: “To judge the likelihood that harm occurred, appellate courts must consider everything in the record including all the evidence admitted at trial, [and] the closing arguments, ... We hold once again that the evidence of the defendant’s guilt is a factor to be considered in any thorough harm analysis.” Id. at 358.
The majority fell into the same trap as the court of appeals in Motilla; its consideration of the weight of the evidence was incomplete. Id. at 358. The opinion glossed over evidence from the guilt/innocence phase and the punishment phase. Granted, we don’t know what might have gone through the jurors’ minds. See Veteto v. State, 8 S.W.3d 805, 813 (Tex.App.Waco 2000, pet. refd). However, the evidence in this case was overwhelming in support of the jury’s assessment of a lengthy prison term rather than probation. Based on this record, I have no trouble concluding beyond a reasonable doubt that the prosecutor’s comments did not contribute to Roberson’s punishment. Tex.R.App. P. 44.2(a). Thus, if the trial court had committed error, that error was harmless.
For the reasons as expressed herein, I respectfully dissent.