While social media data is a valuable source for inferring human behavior, its in-practice utility hinges on extraneous factors. Notable is the “observer effect,” where awareness of being monitored can alter people’s social media use. We present a causal-inference study to examine this phenomenon on the longitudinal Facebook use of 300+ participants who voluntarily shared their data spanning an average of 82 months before and 5 months after study enrollment. We measured deviation from participants’ expected social media use through time series analyses. Individuals with high cognitive ability and low neuroticism decreased posting immediately after enrollment, and those with high openness increased posting. The sharing of self-focused content decreased, while diverse topics emerged. We situate the findings within theories of self-presentation and self-consciousness. We discuss the implications of correcting observer effect in social media data-driven measurements, and how this phenomenon shines light on the ethics of these measurements.

Abstract. A graph [Formula: see text] is Ramsey for a pair of graphs [Formula: see text] if any red/blue-coloring of the edges of [Formula: see text] yields a copy of [Formula: see text] with all edges colored red or a copy of [Formula: see text] with all edges colored blue. Two pairs of graphs are called Ramsey equivalent if they have the same collection of Ramsey graphs. The symmetric setting, that is, the case [Formula: see text], received considerable attention. This led to the open question whether there are connected graphs [Formula: see text] and [Formula: see text] such that [Formula: see text] and [Formula: see text] are Ramsey equivalent. We make progress on the asymmetric version of this question and identify several nontrivial families of Ramsey equivalent pairs of connected graphs. Certain pairs of stars provide a first, albeit trivial, example of Ramsey equivalent pairs of connected graphs. Our first result characterizes all Ramsey equivalent pairs of stars. The rest of the paper focuses on pairs of the form [Formula: see text], where [Formula: see text] is a tree and [Formula: see text] is a complete graph. We show that if [Formula: see text] belongs to a certain family of trees, including all nontrivial stars, then [Formula: see text] is Ramsey equivalent to a family of pairs of the form [Formula: see text], where [Formula: see text] is obtained from [Formula: see text] by attaching disjoint smaller cliques to some of its vertices. In addition, we establish that for [Formula: see text] to be Ramsey equivalent to [Formula: see text], [Formula: see text] must have roughly this form. On the other hand, we prove that for many other trees [Formula: see text], including all odd-diameter trees, [Formula: see text] is not equivalent to any such pair, not even to the pair [Formula: see text], where [Formula: see text] is a complete graph [Formula: see text] with a single edge attached.

Abstract. A graph [Formula: see text] is said to be [Formula: see text]-Ramsey for a [Formula: see text]-tuple of graphs [Formula: see text], denoted by [Formula: see text], if every [Formula: see text]-edge-coloring of [Formula: see text] contains a monochromatic copy of [Formula: see text] in color [Formula: see text] for some [Formula: see text]. Let [Formula: see text] denote the smallest minimum degree of [Formula: see text] over all graphs [Formula: see text] that are minimal [Formula: see text]-Ramsey for [Formula: see text] (with respect to subgraph inclusion). The study of this parameter was initiated in 1976 by Burr, Erdős, and Lovász, who determined its value precisely for a pair of cliques. Over the past two decades the parameter [Formula: see text] has been studied by several groups of authors, their main focus being on the symmetric case, where [Formula: see text] for all [Formula: see text]. The asymmetric case, in contrast, has received much less attention. In this paper, we make progress in this direction, studying asymmetric tuples consisting of cliques, cycles, and trees. We determine [Formula: see text] when [Formula: see text] is a pair of one clique and one tree, a pair of one clique and one cycle, and a pair of two different cycles. We also generalize our results to multiple colors and obtain bounds on [Formula: see text] in terms of the size of the cliques [Formula: see text], the number of cycles, and the number of cliques. Our bounds are tight up to logarithmic factors when two of the three parameters are fixed.

Given any graph $H$, a graph $G$ is said to be $q$-Ramsey for $H$ if every coloring of the edges of $G$ with $q$ colors yields a monochromatic subgraph isomorphic to $H$. Such a graph $G$ is said to be minimal $q$-Ramsey for $H$ if additionally no proper subgraph $G'$ of $G$ is $q$-Ramsey for $H$. In 1976, Burr, Erdös, and Lovász initiated the study of the parameter $s_q(H)$, defined as the smallest minimum degree among all minimal $q$-Ramsey graphs for $H$. In this paper, we consider the problem of determining how many vertices of degree $s_q(H)$ a minimal $q$-Ramsey graph for $H$ can contain. Specifically, we seek to identify graphs for which a minimal $q$-Ramsey graph can contain arbitrarily many such vertices. We call a graph satisfying this property $s_q$-abundant. Among other results, we prove that every cycle is $s_q$-abundant for any integer $q\geq 2$. We also discuss the cases when $H$ is a clique or a clique with a pendant edge, extending previous results of Burr and co-authors and Fox and co-authors. To prove our results and construct suitable minimal Ramsey graphs, we use gadget graphs, which we call pattern gadgets and which generalize earlier constructions used in the study of minimal Ramsey graphs. We provide a new, more constructive proof of the existence of these gadgets.

Given r families of subsets of a fixed n-set, we say that they are r-cross t-intersecting if for every choice of representatives, exactly one from each family, the common intersection of these representatives is of size at least t. We obtain a generalisation of a result by Hilton and Milner on cross intersecting families. In particular, we determine the maximum possible sum of the sizes of non-empty r-cross t-intersecting families in the case when all families are k-uniform and in the case when they are arbitrary subfamilies of the power set. Only some special cases of these results had been proved before. The method we use also yields more general results concerning measures of families instead of their sizes.

Effective ways to measure employee job satisfaction are fraught with problems of scale, misrepresentation, and timeliness. Current methodologies are limited in capturing subjective differences in expectations, needs, and values at work, and they do not lay emphasis on demographic differences, which may impact people's perceptions of job satisfaction. This study proposes an approach to assess job satisfaction by leveraging large-scale social media data. Starting with an initial Twitter dataset of 1.5M posts, we examine two facets of job satisfaction, pay and supervision. By adopting a theory-driven approach, we first build machine learning classifiers to assess perceived job satisfaction with an average AUC of 0.84. We then study demographic differences in perceived job satisfaction by geography, sex, and race in the U.S. For geography, we find that job satisfaction on Twitter exhibits insightful relationships with macroeconomic indicators such as financial wellbeing and unemployment rates. For sex and race, we find that females express greater pay satisfaction but lower supervision satisfaction than males, whereas Whites express the least pay and supervision satisfaction. Unpacking linguistic differences, we find contrasts in different groups' underlying priorities and concerns, e.g., under-represented groups saliently express about basic livelihood, whereas the majority groups saliently express about self-actualization. We discuss the role of frame of reference and the "job satisfaction paradox", conceptualized by organizational psychologists, in explaining our observed differences. We conclude with theoretical and sociotechnical implications of our work for understanding and improving worker wellbeing.