Author Archives: Ian A. Silver

Introduction (PDF) A modifier is a value we identify – either systematically or randomly – that is used to specify the relationship between two variables. Alongside the distributional commands, we can not conduct a comprehensive data simulation without modifiers. We can simply save a value as a letter or use the value independent of aContinue reading “Modifiers”

The results of the looped simulations were updated on 04/04/21. An oversight occurred in the initial looped simulation where all of the sample sizes equaled 10,000 cases. The results did not change by permitting the sample size to vary between 100 and 1,000. The interpretations made when the association between X and Y was specifiedContinue reading “Entry 7: The Exclusion of Confounders (Confounder Bias)”

Introduction (PDF & R-Code) Consistent with the goals of the series, we will discuss how to specify data simulations satisfying the assumptions – rules – that we want to exist within the data. Nevertheless, before we being defining relationships and constructing dataframes, we must learn how to specify a vector of values (e.g, a variable)Continue reading “Simulating Distributions & Variables In R”

Introduction (PDF & R-Code) I apologize; In previous entry I used the term balance to refer to the similarities between the treatment and control group on the key covariates before and after the matching procedure was conducted. By rule of thumb, balance is used to describe the similarities between the treatment and control groups afterContinue reading “Entry X-2: Limited Common Support (Propensity Score Matching)”

For the eight misidentification, we regressed X1 on EX31-EX33, LEN1-LEN4 on X1, and Y1 on X1, LEN1-LEN4, and EX1-EX20 using Lavaan (illustrated in the Figure below). This was replicated 10,000 times – using R-loops – while randomly varying the effects for each causal pathway between the constructs in the network excluding the direct effect of X1 on Y1 (true direct causalContinue reading “Misidentification 8: The Right Estimate”

For the seventh misidentification, we regressed X1 on EX31-EX33, LEN1-LEN4 on X1, and Y1 on X1 and LEN1-LEN4 using Lavaan (illustrated in the Figure below). This was replicated 10,000 times – using R-loops – while randomly varying the effects for each causal pathway between the constructs in the network excluding the direct effect of X1 on Y1 (trueContinue reading “Misidentification 7: Previous Post Plus LEN1-LEN4 Regressed on X1”

Introduction (PDF & R-Code) Conducting a randomized controlled trial is the foremost strategy for generating causal inferences in the social sciences. The required random assignment of the treatment condition, however, limits our ability to use experimental designs when conducting research with ethical concerns. Specifically, we can not randomly assign participants to be born into anContinue reading “Entry X-1: Covariate Imbalance (Propensity Score Matching)”

For the fifth misidentification, we regressed X1 on EX31-EX33, Y1 on X1, EN7-EN12 on X1, covaried the residuals of Y1 and EN7-EN12 using Lavaan (illustrated in the Figure below). This was replicated 10,000 times – using R-loops – while randomly varying the effects for each causal pathway between the constructs in the network excluding the direct effectContinue reading “Misidentification 5: Previous entry plus X1 regressed on EX31-EX33”

For the forth misidentification, we regressed Y1 on X1, regressed EN7-EN12 on X1, and covaried the residuals of Y1 and EN7-EN12 using Lavaan (illustrated in the Figure below). This was replicated 10,000 times – using R-loops – while randomly varying the effects for each causal pathway between the constructs in the network excluding the directContinue reading “Misidentification 4: Y1 regressed on X1, EN7-EN12 regressed on X1, and Y1 covarying with EN7-EN12”