Variable labels are presented for each Excell sheet tab representing a set of analyses reported in the paper. There are a few missing data points for phenotype data. Group Choice Female Perspective (Table 2, Figure 2) Diet : 1= LO diet; 2 = HI diet ID = female ID code Proppref = proportion of her trials in which female was chosen as mate Mass (measured in grams) HW = headwidth (mm) TSNO = number of tail stripe feathers TranTSNO = TSNO2 Beakscore = beak color (Munsell Book of Color) Tranbeak = 1/beakscore2 Age = female’s age at the time of her first trial Group Choice Male Perspective (Figure 3) ID = male ID code Propprefhi = proportion of trials in which male chose a HI-diet female Mass (measured in grams) HW = headwidth (mm) TSNO = number of tail stripe feathers Beakscore = beak color (Munsell Book of Color) Age = male’s age at the time of his first trial Fecundity trials (Table 3, Figure 4) Diet: 1= LO diet; 2 = HI diet ID = female ID code B4eggmass = mean egg mass before LO-female diet switch B4CS = mean clutch size before LO-female diet switch B4Fecundity = B4eggmass x B4CS Aftereggmass = mean egg mass after LO-female diet switch AfterCS = mean clutch size after LO-female diet switch AfterFecundity = Aftereggmass x AfterCS Mass = mass when female assigned to fecundity trial TSNO = Number of tail stripe featherss TranTSNO = TSNO2 HW = headwidth (mm) Beakscore = beak color (Munsell Book of Color) Proppref = proportion of trials in which female was chosen in Group Choice trials (not all Fecundity females participated in those trials, as described in Methods) Age = age at time of first “B4” clutch Female Competition (Table 4) Diet: 1= LO diet; 2 = HI diet ID = female ID code Mass (measured in grams) HW = headwidth (mm) TSNO = number of tail stripe feathers Beakscore = beak color (Munsell Book of Color) Age = female’s age at the time she entered Competition trial Behaviors (Before and After males were added to trial): Ag = mean incidence of female initiating aggressive interactions with other females TranAg = square root (Ag) Agby = mean incidence of female being recipient of aggressive interactions from other females Tranagby = square root (Agby) DP [Dyadic Preference] Female Phenotype (Text only) Diet : 1= LO diet; 2 = HI diet ID = female ID code Mass (measured in grams) HW = headwidth (mm) TranHW = 1/HW3 TSNO = number of tail stripe feathers TranTSNO = TSNO2 Beakscore = beak color (Munsell Book of Color) Tranbeak = Beakscore3 Age = female’s age at the time of her first trial Propprefaudab = Mean proportion of social time with chooser males when audience was absent Propprefaudpres = Mean proportion of social time with chooser males when audience was present DP Female Affiliation (Table 5) Diet : 1= LO diet; 2 = HI diet ID = female ID code Proptime = Proportion of male’s social time a female received in a given trial Totaffbeh = Number of Affiliative behaviors female had with male in a given trial Trantotaffbeh = square root (Totaffbeh) DP Male Perspective (Influence of Test Order/Audience Effect) – Figure 5 and Figure 6; Tables 6,7, 8 ID = male ID Mass (measured in grams) HW = headwidth (mm) TSNO = number of tail stripe feathers TranTSNO = TSNO2 Beakscore = beak color (Munsell Book of Color) Testtype: 1 = audience males absent ; 2 = audience males present Testorder: 1 = audience males absent; 2 = audience males present Sumtime = total social time spent with stimulus females (minutes) Transumtime = sumtime12/(1023) Propprefhi = proportion of social time spent with HI-diet females

Understanding the dynamics of mutual mate choice requires investigation of mate preferences of both sexes using a variety of designs, but fewer studies have focused on male choice in avian models. Here we conducted two experiments on preferences of male zebra finches to study the impact of trial design on results. Experimental design varied in number of trial participants, inclusion of observer (“audience”) males, and housing design for stimulus females. Females were reared on one of two diets to enhance variation in mating quality: those reared on a protein-supplement diet (HI-diet) were predicted to be more attractive to males than (LO-diet) females that did not receive supplements. Results differed among trial types. Notably, males showed the predicted preference for HI-diet females only in the Group Choice experiment, where two male subjects simultaneously chose from a field of four females, and all six birds interacted freely. In the Dyadic Preference (DP) experiment, a single male was allowed to interact with two stimulus females that were physically isolated; in half of these trials, audience males were present. In DP trials without audience males, test males did not express a preference consensus; however, with audience males present, test males preferred LO-diet females. Results are consistent with a small but growing literature indicating that results of mate choice experiments can be highly sensitive to design considerations.

Data were collected in the course of laboratory experiments, archived by the first author, and statistically analyzed by both authors. Analyses were performed in STATA 15.0.