Article Figures & Data
Figures
- Figure 1.
Judges' decisions to invalidate a will based on undue influence.
- Figure 2.
Differences in factor ratings (confidence intervals) between judges who did versus did not identify undue influence in Case 1. Positive values reflect factors rated as more influential by judges who found that undue influence was present.
Tables
Case 1: John Doe died of cancer at 58 years of age, leaving his estate to Jane Roe, whom he refers to in his will as “my wife.” The couple had exchanged vows in another country and were cohabiting at the time of Mr. Doe's death, as they had for most of their relationship. When their relationship began, Mr. Doe had been widowed for four years, and Ms. Roe was married and still living with her seventh husband. She moved in with Mr. Doe shortly after meeting him and later obtained a divorce from her husband. Mr. Doe's two adult sons have contested the will, alleging undue influence on the part of Ms. Roe. The sons believe that Mr. Doe was afraid of Ms. Roe and claim the couple were both alcoholics and argued frequently. Mr. Doe's physician opines that Mr. Doe's cancer and other physical health problems could have had an effect on his judgment and personality. Mr. Doe's will was executed approximately a year before his death. Case 2: Jane Doe, a 78-year-old disabled nursing home resident, left the bulk of her estate to Mr. H., a 35-year-old male social worker who worked with and visited her regularly as she neared death. Ms. Doe's only living relatives, a brother and niece, contested the will, alleging undue influence on the part of Mr. H. Ms. Doe had never married. Ms. Doe and Mr. H. first met when Ms. Doe was admitted to the nursing home with terminal lung cancer. In accordance with his social work duties, Mr. H. took control of Ms. Doe's finances, as she was physically unable to manage them on her own. During her acquaintance with Mr. H., Ms. Doe executed her first and only will, leaving most of her estate to Mr. H. and a modest life estate to her brother. Ms. Doe had said that she loved Mr. H. and informed her attorney that Mr. H. was her best friend. Ms. Doe executed her will approximately nine months before her death. The cases in Survey Version B are identical except that the genders of testators and beneficiaries are reversed (see Table 2).
Version A Version B Case 1 Male testator (female beneficiary) Female testator (male beneficiary) Case 2 Female testator (male beneficiary) Male testator (female beneficiary) Variable All (n = 117) Survey Version A (n = 68) Survey Version B (n = 49) χ2 (df)* p Female 23.3 19.4 28.6 1.33 (1) .25 Mean age, years 4.32 (2) .12 <50 21.4 19.1 24.5 51–60 47.9 55.9 36.7 >60 30.8 25 38.8 Judicial experience, years .26 (2) .88 <5 27.4 27.9 26.5 6–10 22.2 23.5 20.4 >10 50.4 48.5 53.1 Will contests presided over involving alleged undue influence 2.51 (3) .47 None 9.5 7.4 12.5 1–5 44 48.5 37.5 6–10 22.4 23.5 20.8 >10 24.1 20.6 29.2 Prior mental health training, yes 28.3 28.4 28.3 0 (1) .99 Survey Version A: Case 1 involved a male testator; Case 2 involved a female testator.
Survey Version B: Case 1 involved a female testator; Case 2 involved a male testator.
Values are percentages; sums may not equal 100% due to rounding.
↵* Respondents to Survey A versus Survey B.
State Probate Judges Invited to Participate From Each State (n) Probate Judges Who Responded (n) Response Rate Per State (%) Proportion of the Total Sample (%) Connecticut 118 21 18 18 Kansas 163 16 10 14 Massachusetts 46 9 20 8 Mississippi 48 10 21 9 Nebraska 55 1 2 1 New York 76 10 13 9 Oklahoma 236 23 10 20 Rhode Island 39 19 49* 16 South Carolina 46 8 17 7 Total 827 117 — 100 ↵* The high response rate for Rhode Island is probably due to the principal investigator and other researchers being located there.
Model A Model B Model C Estimate Standard error p Estimate Standard error p Estimate Standard error p Intercept 0.975 0.398 .014 −2.120 0.380 .000 0.340 0.586 .566 Case −2.192 0.335 .000 2.224 0.341 .000 −2.041 0.966 .035 Testator gender (female vs. male) −0.141 0.325 .666 0.522 0.380 .169 0.868 0.881 .325 Judge gender (female vs. male) 0.894 0.386 .021 1.718 0.573 .003 1.678 0.791 .034 Testator gender: judge Gender −1.569 0.779 .044 −1.115 1.108 .314 Case: testator gender 0.278 1.327 .834 Case: judge gender −0.034 1.157 .976 Case: testator gender: judge gender −0.694 1.557 .656 A generalized linear mixed-effects regression model with random intercepts was used for repeated assessments. Model A includes main effects only. Model B adds a planned, two-way interaction of testator gender by judge gender. Model C adds an additional evaluation of the three-way interaction along with all lower order terms.
