Purrble RCT Full Analyses: Descriptives and Preliminary Results

Recording Keeping: There are two master files that we are using for analyses. They are essentially the same file, though one is in wide format and the other is in long format.

The wide format dataset is called “Purrble_Wide_Final.” The long dataset format dataset is called “NoDup_PurrbleAnon.”

This dataset includes the N=154 participants who were included in the randomized control trial examining Purrble with a population of university students, with an emphasis on the LGTBQ+ community.

Participation in Each Week over Time Note: Weeks 1-3 were considered “pre-test.” Purrble was given (or not) after week 3. Weeks 11-13 are considered “Post-test”.

### **Number of Participants in Each Condition**

Participant Counts by Condition

Condition	N
Purrble	76
Waitlist Control	78

### **Completion Counts Over Time**

Number of Participants Completing Each Week

Week	Count
1	147
2	149
3	150
4	141
5	139
6	139
7	139
8	142
9	127
10	128
11	128
12	118
13	131

Attrition Analysis by Condition Attrition is defined as dropping out of study prior to Week 11. We checked to examine whether there were differences in drop out for the Waitlist and Treatment groups. Result: No difference in attrition by group

### **Attrition Analysis by Condition (Prior to Week 11)**

Descriptive Statistics for Attrition by Condition

condition	N	Mean	SD
Purrble Treatment	76	0.0921053	0.2910959
Waitlist Control	78	0.0641026	0.2465209

### **T-Test Results**

T-Test Comparing Attrition by Condition

Mean (Condition 1)	Mean (Condition 2)	t-value	p-value	Lower CI	Upper CI
0.0921053	0.0641026	0.6434268	0.5209519	-0.0580069	0.1140123

Demographics Descriptive Statistics by Condition

Gender and Sexual Identities by Condition

Frequency Table of Gender Identity by Condition

gi_simplified	Purrble	Waitlist Control
Cisgender	39	37
Transgender/genderdiverse	37	40
NA	0	1

Frequency Table of Sexual Orientation by Condition

so_simplified	Purrble	Waitlist Control
Asexual	13	9
Bisexual	28	25
Demisexual	2	1
Gay/lesbian	11	18
Heterosexual	1	0
Pansexual	8	10
Queer	13	15

Race and Nationality by Condition*

Frequency Table of Race Variables by Condition

Race_Variable	Purrble	Waitlist Control
Race_Asian	10	17
Race_Arabic	0	1
Race_Black	1	3
Race_Hispanic	2	0
Race_Pacific	0	0
Race_White	59	55
Race_unknown	9	6

4 people in the Purrble condition reported multiple racial/ethnic identities
4 people in the Waitlist Control condition reported multiple racial/ethnic identities

Frequency Table of Nationality by Condition

Nationality	Purrble	Waitlist Control
British	35	36
British-Carribean	1	1
British-Indian	1	0
Chinese	1	5
Filipino	1	0
Indian	3	5
Irish	1	1
Irish-American	1	0
Mexican	1	0
NR	28	21
Pakistani	1	0
Polish	2	2
Bangladeshi	0	1
British-Japanese	0	1
British-Pakistani	0	1
Indonesian	0	1
Iranian	0	1
Irish-Carribean	0	1
Malaysian Chinese	0	1

Age by Condition The table below presents age descriptives by condition. A t-test was conducted to determine whether there were baseline differences in age. There were no significant differences between the two conditions.

### **Age by Condition (Baseline)**

Descriptive Statistics for Age by Condition

	Condition	N	Mean	SD	Min	Max
X11	Purrble	76	20.40789	2.281389	16	25
X12	Waitlist Control	78	20.06410	2.456583	16	25

### **T-test for Age by Condition**

T-test Results for Age by Condition

Mean (Waitlist Control)	Mean (Purrble)	t-value	p-value	Lower CI	Upper CI
20.40789	20.0641	0.9001998	0.3694412	-0.410752	1.098336

### **Cohen's d for Age by Condition**

Effect Size (Cohen’s d) for Age by Condition

Cohen’s d	95% CI Lower	95% CI Upper
0.1449526	-0.1716078	0.4610376

Preliminary Analyses First, we want to check to see if there are any differences in the outcome(s) of interest at baseline, by chance Result:* No difference in baseline pre-test measures- woo!

Baseline differences in study outcomes by condition

### **Pre-Test Descriptive Statistics**

Descriptive Statistics for Pre-Test Data

	N	Mean	SD	Min	Max	Skewness	Kurtosis
Pre_DERS8_Sum	153	28.136	4.705	14.333	38.333	-0.413	-0.122
Pre_GAD7_Sum	153	13.715	3.976	3.000	22.000	-0.167	-0.441
Pre_PHQ9_Sum	153	15.031	4.569	3.000	26.667	-0.011	-0.086
Pre_SHS_Pathways	149	13.292	4.284	3.000	24.000	-0.136	-0.403
Pre_SHS_Agency	149	10.708	4.929	3.000	24.000	0.339	-0.645
Pre_SHS_TotalHope	149	24.000	8.325	8.000	46.000	0.282	-0.289
Pre_ucla_Sum	148	7.095	1.618	3.000	9.000	-0.506	-0.662
Pre_pmerq_Focus_Avg	149	2.737	1.059	1.000	6.000	0.421	-0.076
Pre_pmerq_Distract_Avg	149	4.222	1.127	1.000	6.000	-0.833	0.616
Pre_pmerq_AD_Avg	149	3.480	0.923	1.000	6.000	-0.320	0.504

### **Post-Test Descriptive Statistics**

Descriptive Statistics for Post-Test Data

	N	Mean	SD	Min	Max	Skewness	Kurtosis
Post_DERS8_Sum	142	26.965	7.318	8	40	-0.264	-0.821
Post_GAD7_Sum	142	12.630	4.980	1	22	-0.080	-0.761
Post_PHQ9_Sum	142	14.326	6.310	0	27	-0.009	-0.682
Post_SHS_Pathways	131	14.664	4.309	3	24	-0.248	-0.455
Post_SHS_Agency	131	12.626	5.213	3	24	-0.004	-0.846
Post_SHS_TotalHope	131	27.290	8.796	6	47	-0.043	-0.488
Post_ucla_Sum	131	6.802	1.703	3	9	-0.415	-0.682
Post_pmerq_Focus_Avg	130	2.992	1.193	1	6	0.282	-0.322
Post_pmerq_Distract_Avg	130	4.337	1.054	1	6	-1.133	1.673
Post_pmerq_AD_Avg	130	3.665	0.951	1	6	-0.316	0.917

### **ANCOVA Results by Outcome**

Analysis of Covariance (ANCOVA) Results by Outcome

Outcome	Pre-Test Covariate	Mean_Purrble	Mean_Waitlist Control	SD_Purrble	SD_Waitlist Control	F-Value	p-Value
Post_DERS8_Sum	Pre_DERS8_Sum	25.261	28.575	7.799	6.481	13.075	0.000
Post_GAD7_Sum	Pre_GAD7_Sum	12.002	13.224	5.465	4.431	3.498	0.064
Post_PHQ9_Sum	Pre_PHQ9_Sum	13.442	15.162	6.658	5.886	6.195	0.014
Post_SHS_Pathways	Pre_SHS_Pathways	14.468	14.841	4.456	4.196	0.642	0.424
Post_SHS_Agency	Pre_SHS_Agency	12.516	12.725	5.416	5.061	0.172	0.679
Post_SHS_TotalHope	Pre_SHS_TotalHope	26.984	27.565	9.035	8.632	0.421	0.518
Post_ucla_Sum	Pre_ucla_Sum	6.677	6.913	1.845	1.569	1.092	0.298
Post_pmerq_Focus_Avg	Pre_pmerq_Focus_Avg	3.105	2.890	1.206	1.181	1.128	0.290
Post_pmerq_Distract_Avg	Pre_pmerq_Distract_Avg	4.419	4.262	1.062	1.049	0.770	0.382
Post_pmerq_AD_Avg	Pre_pmerq_AD_Avg	3.762	3.576	0.932	0.967	1.492	0.224

Follow-up: Effect Sizes for Significant/Marginal Variables (I’m sorry I don’t know how to make this look pretty! If anyone knows please help!) Cohen’s F For Each Significant/Marginal Outcome: Anxiety: . 0.16 Depression: 0.21 DERS: 0.31

---------------------------------------
ANCOVA for: Post_DERS8_Sum controlling for Pre_DERS8_Sum 
---------------------------------------
               Df Sum Sq Mean Sq F value   Pr(>F)    
condition       1    410   410.4   13.07 0.000419 ***
Pre_DERS8_Sum   1   2779  2778.5   88.53  < 2e-16 ***
Residuals     138   4331    31.4                     
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
13 observations deleted due to missingness

Partial Eta Squared:
# Effect Size for ANOVA (Type I)

Parameter     | Eta2 (partial) |       95% CI
---------------------------------------------
condition     |           0.09 | [0.03, 1.00]
Pre_DERS8_Sum |           0.39 | [0.29, 1.00]

- One-sided CIs: upper bound fixed at [1.00].
Cohen's f (Partial):
# Effect Size for ANOVA (Type I)

Parameter     | Cohen's f (partial) |      95% CI
-------------------------------------------------
condition     |                0.31 | [0.16, Inf]
Pre_DERS8_Sum |                0.80 | [0.64, Inf]

- One-sided CIs: upper bound fixed at [Inf].
---------------------------------------
ANCOVA for: Post_GAD7_Sum controlling for Pre_GAD7_Sum 
---------------------------------------
              Df Sum Sq Mean Sq F value   Pr(>F)    
condition      1     54    54.0   3.498   0.0636 .  
Pre_GAD7_Sum   1   1311  1311.2  84.884 4.68e-16 ***
Residuals    138   2132    15.4                     
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
13 observations deleted due to missingness

Partial Eta Squared:
# Effect Size for ANOVA (Type I)

Parameter    | Eta2 (partial) |       95% CI
--------------------------------------------
condition    |           0.02 | [0.00, 1.00]
Pre_GAD7_Sum |           0.38 | [0.28, 1.00]

- One-sided CIs: upper bound fixed at [1.00].
Cohen's f (Partial):
# Effect Size for ANOVA (Type I)

Parameter    | Cohen's f (partial) |      95% CI
------------------------------------------------
condition    |                0.16 | [0.00, Inf]
Pre_GAD7_Sum |                0.78 | [0.62, Inf]

- One-sided CIs: upper bound fixed at [Inf].
---------------------------------------
ANCOVA for: Post_PHQ9_Sum controlling for Pre_PHQ9_Sum 
---------------------------------------
              Df Sum Sq Mean Sq F value Pr(>F)    
condition      1  110.5   110.5   6.195  0.014 *  
Pre_PHQ9_Sum   1 3034.1  3034.1 170.110 <2e-16 ***
Residuals    138 2461.4    17.8                   
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
13 observations deleted due to missingness

Partial Eta Squared:
# Effect Size for ANOVA (Type I)

Parameter    | Eta2 (partial) |       95% CI
--------------------------------------------
condition    |           0.04 | [0.00, 1.00]
Pre_PHQ9_Sum |           0.55 | [0.46, 1.00]

- One-sided CIs: upper bound fixed at [1.00].
Cohen's f (Partial):
# Effect Size for ANOVA (Type I)

Parameter    | Cohen's f (partial) |      95% CI
------------------------------------------------
condition    |                0.21 | [0.07, Inf]
Pre_PHQ9_Sum |                1.11 | [0.93, Inf]

- One-sided CIs: upper bound fixed at [Inf].

Self-harm Questions

Post-test Logistic Regression to Investigate Intervention Effects on Self-Harm Outcomes Result: Condition was not a significant predictor of any self-harm outcome (coded binary).

Characteristic	SHQ1 Model		SHQ2 Model		SHQ3 Model		SHQ_Any Model
	OR1,2	SE2	OR1,2	SE2	OR1,2	SE2	OR1,2	SE2
condition
Purrble Treatment	—	—	—	—	—	—	—	—
Waitlist Control	0.87	0.452	1.02	0.412	1.15	0.546	0.91	0.434
SHQ1_2	11.6***	0.484
SHQ2_2			4.36***	0.408
SHQ3_2					3.14*	0.559
SHQ_Any_2							5.83***	0.486
1 *p<0.05; **p<0.01; ***p<0.001
2 OR = Odds Ratio, SE = Standard Error

Below, we display a table and graph of the frequency of responses for all self-harm questions. Please note- they are all equal (which is why the lines all overlap and there appears to only be one!)

Number of Responses for Self-Harm Questions Over Time
Week	N_SHQ1	N_SHQ2	N_SHQ3
1	147	147	147
2	149	149	149
3	150	150	150
4	141	141	141
5	140	140	140
6	139	139	139
7	141	141	141
8	142	142	142
9	128	128	128
10	128	128	128
11	129	129	129
12	118	118	118
13	131	131	131

Linear Mixed Effects Models Results interpretation:

Overall Summary of Significant Results

• Purrble showed greater improvement compared to the Waitlist group for:
  • Emotional Dysregulation (DERS8)
  • Anxiety (GAD7)
  • Depression (PHQ9)
  • Positive Emotion Regulation - Focus (pmerq_Focus_Avg)
  • Positive Emotion Regulation - Attentional Deployment (pmerq_AD_Avg)

---

Fixed Effects Results

1. Emotional Dysregulation (DERS8_Sum)

• Interaction Effect: The Waitlist group showed a significantly less steep decline than Purrble (p < 0.001).
• Interpretation: Emotional dysregulation improved more in the Purrble group compared to Waitlist.

2. Anxiety Symptoms (GAD7_Sum)

• Interaction Effect: The Waitlist group experienced a significantly smaller improvement than Purrble (p = 0.016).
• Interpretation: Anxiety symptoms decreased more in the Purrble group than in Waitlist.

3. Depression Symptoms (PHQ9_Sum)

• Interaction Effect: The Waitlist group showed significantly smaller improvement than Purrble (p < 0.001).
• Interpretation: The Purrble group experienced a greater reduction in depressive symptoms.

4. Positive Emotion Regulation - Focus (pmerq_Focus_Avg)

• Interaction Effect: The Purrble group showed a slightly greater improvement than the Waitlist group (p = 0.018).
• Interpretation: The Purrble group had enhanced emotion regulation strategies focused on positive emotions.

5. Positive Emotion Regulation - Attentional Deployment (pmerq_AD_Avg)

• Interaction Effect: The Purrble group showed significantly greater improvement compared to the Waitlist group (p = 0.010).
• Interpretation: The Purrble group improved more in acceptance and dampening strategies.

---

Non-Significant Interaction Effects

For the following variables, there was no significant difference between groups over time: - Hope Pathways (SHS_Pathways) - Hope Agency (SHS_Agency) - Total Hope (SHS_TotalHope) - Loneliness (UCLA_Sum) - Positive Emotion Regulation - Distraction (pmerq_Distract_Avg) (marginal trend, p = 0.063)

-------------------------------------------------
Fixed Effects for Outcome: DERS8_Sum 
-------------------------------------------------

Estimate	Std. Error	df	t value	Pr(>|t|)	Variable
28.383	0.674	203.773	42.116	0.000	(Intercept)
-0.283	0.038	1631.002	-7.502	0.000	Week
-0.332	0.947	203.540	-0.351	0.726	conditionWaitlist Control
0.296	0.052	1630.582	5.669	0.000	Week:conditionWaitlist Control

-------------------------------------------------
Fixed Effects for Outcome: GAD7_Sum 
-------------------------------------------------

Estimate	Std. Error	df	t value	Pr(>|t|)	Variable
13.779	0.497	225.662	27.749	0.000	(Intercept)
-0.161	0.032	1633.666	-5.044	0.000	Week
-0.139	0.697	225.254	-0.199	0.842	conditionWaitlist Control
0.107	0.044	1632.714	2.411	0.016	Week:conditionWaitlist Control

-------------------------------------------------
Fixed Effects for Outcome: PHQ9_Sum 
-------------------------------------------------

Estimate	Std. Error	df	t value	Pr(>|t|)	Variable
15.778	0.606	198.061	26.027	0.000	(Intercept)
-0.190	0.033	1629.309	-5.835	0.000	Week
-1.288	0.852	197.807	-1.512	0.132	conditionWaitlist Control
0.233	0.045	1628.800	5.182	0.000	Week:conditionWaitlist Control

-------------------------------------------------
Fixed Effects for Outcome: SHS_Pathways 
-------------------------------------------------

Estimate	Std. Error	df	t value	Pr(>|t|)	Variable
12.271	0.566	351.869	21.692	0.000	(Intercept)
0.178	0.050	274.989	3.587	0.000	Week
0.972	0.799	354.277	1.217	0.224	conditionWaitlist Control
-0.066	0.070	273.258	-0.946	0.345	Week:conditionWaitlist Control

-------------------------------------------------
Fixed Effects for Outcome: SHS_Agency 
-------------------------------------------------

Estimate	Std. Error	df	t value	Pr(>|t|)	Variable
9.352	0.651	349.587	14.375	0.000	(Intercept)
0.241	0.057	275.121	4.234	0.000	Week
1.060	0.918	352.019	1.154	0.249	conditionWaitlist Control
-0.079	0.080	273.408	-0.999	0.319	Week:conditionWaitlist Control

-------------------------------------------------
Fixed Effects for Outcome: SHS_TotalHope 
-------------------------------------------------

Estimate	Std. Error	df	t value	Pr(>|t|)	Variable
21.619	1.090	334.610	19.833	0.000	(Intercept)
0.422	0.092	273.920	4.570	0.000	Week
2.025	1.538	337.177	1.316	0.189	conditionWaitlist Control
-0.147	0.129	272.298	-1.139	0.256	Week:conditionWaitlist Control

-------------------------------------------------
Fixed Effects for Outcome: ucla_Sum 
-------------------------------------------------

Estimate	Std. Error	df	t value	Pr(>|t|)	Variable
7.024	0.214	285.701	32.818	0.000	(Intercept)
-0.028	0.016	269.257	-1.728	0.085	Week
0.322	0.301	287.117	1.069	0.286	conditionWaitlist Control
-0.007	0.022	267.706	-0.333	0.740	Week:conditionWaitlist Control

-------------------------------------------------
Fixed Effects for Outcome: pmerq_Focus_Avg 
-------------------------------------------------

Estimate	Std. Error	df	t value	Pr(>|t|)	Variable
2.512	0.143	306.792	17.566	0.000	(Intercept)
0.047	0.011	270.346	4.159	0.000	Week
0.276	0.202	309.628	1.370	0.172	conditionWaitlist Control
-0.038	0.016	269.074	-2.378	0.018	Week:conditionWaitlist Control

-------------------------------------------------
Fixed Effects for Outcome: pmerq_Distract_Avg 
-------------------------------------------------

Estimate	Std. Error	df	t value	Pr(>|t|)	Variable
4.044	0.142	345.894	28.434	0.000	(Intercept)
0.031	0.012	273.095	2.521	0.012	Week
0.251	0.201	348.557	1.249	0.212	conditionWaitlist Control
-0.032	0.017	271.632	-1.864	0.063	Week:conditionWaitlist Control

-------------------------------------------------
Fixed Effects for Outcome: pmerq_AD_Avg 
-------------------------------------------------

Estimate	Std. Error	df	t value	Pr(>|t|)	Variable
3.278	0.120	311.891	27.366	0.00	(Intercept)
0.039	0.010	271.195	4.094	0.00	Week
0.264	0.169	314.726	1.559	0.12	conditionWaitlist Control
-0.035	0.013	269.901	-2.603	0.01	Week:conditionWaitlist Control