How self-awareness is connected to less experience of action crises in personal goal pursuit

In this research, we applied a differential perspective to the study of action crises, i.e., being in an intra-psychic decisional conflict whether to pursue or abandon a goal once difficulties in its pursuit arise. In two studies, we investigated the role of individuals’ levels of self-awareness when experiencing such action crises. Both among professional ballet dancers (daily diary, Study 1) and university undergraduates (preregistered experience sampling, Study 2), individuals with greater levels of (dispositional and situational) self-awareness showed an adaptive, that is, problem-solving oriented way of dealing with difficulties in the pursuit of their (training or study) goals. As a consequence, self-awareness contributed to less experience of action crisis during goal pursuit and, as a result, led to better goal performance. Supplementary Information The online version contains supplementary material available at 10.1007/s11031-022-09942-5.

questionnaires, which took ca. 2 minutes and had to be answered on a smart phone within 30 minutes before the link expired. Between consecutive messages there was always a time lag of at least 40 min.
We used Survey Signal, an online service for experience sampling studies, to send all text messages accordingly. 2 days prior to the start of the experience sampling phase, participants received an initial text message containing a hyperlink to confirm their mobile number. They had to confirm their mobile until midnight before the start of the experience sampling phase.
Due to a technical failure on the second day of the experience sampling phase, only three out of five messages were sent on that day. Thus reducing the number of brief questionnaires to 48.
Participants had to answer at least 38 (80%) of the brief questionnaires sent during the two weeks. Participants who had a lower response rate (but at least 28 responses) received another 10 brief surveys during the next two workdays. We monitored response rates during the experience sampling phase and contacted participants via text message if they did not answer any brief questionnaires or only very few. In order to save costs, we continuously excluded participants who skipped 19 questionnaires and/or answered that they wanted to quit.
Participants who did not reach the 80% response rate even after extended data collection were excluded from further participation (i.e., the follow-up survey). Participants who reached a 90% response rate overall received a bonus payment (CHF 20) in addition to study compensation. Participants who were excluded due to technical problems received partial compensation.
Data of participants who voluntarily dropped out or were excluded during the experience sampling phase was used for analysis unless participants asked for their data to be deleted.

Participant dropout and exclusions:
Suppl. To ensure that brief questionnaires were not only skimmed through, we excluded responses that were completed within less than 20 seconds (15 seconds for evening Supplemental Online Material to SELF-AWARENESS AND ACTION CRISES questionnaires). We also intended to exclude any responses that took more than 30 minutes but none did.
Among the five brief questionnaires sent out every day, the fifth one (evening questionnaire) did not contain all relevant variables and is therefore not used for the present analyses. Unfortunately, while monitoring the data collection, we could not distinguish between complete and partially answered questionnaires. Therefore, a relevant proportion of our data has missing values for one or several of the relevant variables and could not be used.
Furthermore, we were only interested in moments when participants were actually engaged in an activity related to their academic goal.

Deviations from preregistered analysis procedure in Study 2:
Suppl. Table 3

Hypothesis
Preregistered method Reported method Rationale Baseline (& follow-up) data: Baseline problem solving orientation mediates the negative association between self-awareness at the baseline and action crisis in the follow-up survey.
Mediation analysis using PROCESS macro for SPSS by Hayes (2001).
Using consistent software for all analyses.

Experience sampling data:
Momentary problem solving orientation mediates the negative association between momentary self-awareness and doubts.
Clustered mediation analysis using PROCESS macro for SPSS by Hayes (2012).

Analysis of data on the level of days:
Experience sampling and problem solving each aggregated across up to 4 questionnaires per day; progress measured in the 5 th (evening) questionnaire.
Multilevel serial mediation analysis using sermedMLM by the second author in R (available on OSF: [link omitted for blind review]).
Used data on the level of brief questionnaires (up to 4 times per day).
Using sermedLM allowed us (1.) to disentangle within-and between person effects and (2.) to extend the mediation model and include progress as a subsequent dependent variable.
Brief questionnaire data allowed more finegrained analysis than aggregated data per day. Furthermore, daily aggregates may include data from varying numbers of brief questionnaires, thus essentially weighting individual responses differentily.

Descriptive Statistics on Study 2:
Suppl.

Extended Results from Study 2:
Suppl. Table 6

: Regression models of the between-persons mediation analyses in Study 2 with (A) and without (B) baseline action crisis as a covariate
Version A: with covariate (reported in Fig. 2