2b:null 2c:["$","title",null,{"children":"PSEO Margins Overview | Jody Hoon-Starr | Observable"}] 31:T59a,The information shown here aims to give an overview on what we're thinking in terms of the future for [PSEO Data](https://lehd.ces.census.gov/data/pseo_experimental.html), [PSEO Explorer](https://lehd.ces.census.gov/applications/pseo/), and future tools. As of this writing most of the focus within the PSEO project has been on releasing college outcomes per partner institution. As the program grows, both in **number of partner institutions** and **number of years covered** within the data, we're focusing on things that could be done better. As the number of institutions grows we're able to start research into producing **"margins"** (also known as aggregates, roll-ups, etc.) which take the existing protected data and allows us to produce *generic outcomes* - outside the scope of a specific institution. This is what will be done in the following examples. One of the primary concerns with creating these margins is being aware of and handling *composition effects*. For example, what happens if we have two institutions in a state - one flagship institution and a smaller local college? If we aggregated these two schools together to paint a picture of higher education within a state would it be representative? If not, perhaps we shouldn't publish a margin for this particular state. This concern is being discussed internally and is *not* addressed in the examples below. It is noted here for posterity and discussion.32:Td9b,single_cip_chart = { // convert the data into a longform and filter only selected year postgrad const data = convertToLongform(earningsMargins).filter( (e) => e.yearPostgrad == yearInput && e.cipcode == "5102" && e.institution_statefips == "00" ); // have this lookup so that the sort order can be defined const rowKeysLookup = new Map( data.map((d) => { const result = generateRowKeyLookup(d); return [result.id, result.label]; }) ); const sortedRowKeys = [...rowKeysLookup.keys()].sort(); const dataGrouped = d3.group(data, (d) => generateRowKeys(d)); const percentiles = new Set(data.map((d) => d.percentile)); // Specify the chart’s dimensions. const width = 928; const height = dataGrouped.size * 46; const marginTop = 30; const marginRight = 100; const marginBottom = 10; const marginLeft = 450; // scale values const min = d3.min(data.map((e) => e.earnings)); const max = d3.max(data.map((e) => e.earnings)); // Prepare the scales for positional and color encodings const x = d3 .scaleLinear() .domain([min, max]) .rangeRound([marginLeft, width - marginRight]) .nice([6]); const [xMin, xMax] = x.domain(); const y = d3 .scalePoint() .domain(sortedRowKeys.map((k) => rowKeysLookup.get(k))) .rangeRound([marginTop, height - marginBottom]) .padding(1); const color = d3 .scaleOrdinal() .domain(percentiles) .range(d3.schemeTableau10) .unknown("#ccc"); // Create the SVG container. const svg = d3 .create("svg") .attr("width", width) .attr("height", height) .attr("viewBox", [0, 0, width, height]) .attr("style", "max-width: 100%; height: auto;"); // Create the x axis. svg .append("g") .attr("transform", `translate(0,${marginTop})`) .call( d3 .axisTop(x) .ticks(7) .tickFormat((d) => "$" + d3.format(".2s")(d)) ) .call((g) => g .append("text") .text("earnings →") .attr("fill", "currentColor") .attr("transform", `translate(${width - marginRight},0)`) .attr("text-anchor", "end") .attr("dy", -22) ) .call((g) => g .selectAll(".tick line") .clone() .attr("stroke-opacity", 0.1) .attr("y2", height - marginBottom) ) .call((g) => g.selectAll(".domain").remove()); // Add a g container for each group const g = svg .append("g") .attr("text-anchor", "end") .style("font", "14px sans-serif") .selectAll() .data(dataGrouped) .join("g") .attr("transform", ([key]) => `translate(0,${y(key)})`); // Append a grey line to each container. g.append("line") .attr("stroke", "#aaa") .attr("stroke-width", 3) .attr("x1", ([, values]) => x(d3.min(values, (d) => d.earnings))) .attr("x2", ([, values]) => x(d3.max(values, (d) => d.earnings))); // Append the dots to each container. g.append("g") .selectAll() .data(([, values]) => values) .join("circle") .attr("cx", (d) => x(d.earnings)) .attr("fill", (d) => color(d.percentile)) .attr("stroke", "white") .attr("stroke-width", "2") .attr("r", 9); // Append the label to each container. g.append("text") .attr("dy", "0.35em") .attr("x", () => x(xMin) - 6) .text(([group, values]) => group); // Expose the color scale for a legend. return Object.assign(svg.node(), { scales: { color } }); }33:Td44,chart = { // convert the data into a longform and filter only selected year postgrad const data = convertToLongform(earningsMargins).filter( (e) => e.yearPostgrad == yearInput ); // have this lookup so that the sort order can be defined const rowKeysLookup = new Map( data.map((d) => { const result = generateRowKeyLookup(d); return [result.id, result.label]; }) ); const sortedRowKeys = [...rowKeysLookup.keys()].sort(); const dataGrouped = d3.group(data, (d) => generateRowKeys(d)); const percentiles = new Set(data.map((d) => d.percentile)); // Specify the chart’s dimensions. const width = 928; const height = dataGrouped.size * 46; const marginTop = 30; const marginRight = 100; const marginBottom = 10; const marginLeft = 450; // scale values const min = d3.min(data.map((e) => e.earnings)); const max = d3.max(data.map((e) => e.earnings)); // Prepare the scales for positional and color encodings const x = d3 .scaleLinear() .domain([min, max]) .rangeRound([marginLeft, width - marginRight]) .nice([6]); const [xMin, xMax] = x.domain(); const y = d3 .scalePoint() .domain(sortedRowKeys.map((k) => rowKeysLookup.get(k))) .rangeRound([marginTop, height - marginBottom]) .padding(1); const color = d3 .scaleOrdinal() .domain(percentiles) .range(d3.schemeTableau10) .unknown("#ccc"); // Create the SVG container. const svg = d3 .create("svg") .attr("width", width) .attr("height", height) .attr("viewBox", [0, 0, width, height]) .attr("style", "max-width: 100%; height: auto;"); // Create the x axis. svg .append("g") .attr("transform", `translate(0,${marginTop})`) .call( d3 .axisTop(x) .ticks(7) .tickFormat((d) => "$" + d3.format(".2s")(d)) ) .call((g) => g .append("text") .text("earnings →") .attr("fill", "currentColor") .attr("transform", `translate(${width - marginRight},0)`) .attr("text-anchor", "end") .attr("dy", -22) ) .call((g) => g .selectAll(".tick line") .clone() .attr("stroke-opacity", 0.1) .attr("y2", height - marginBottom) ) .call((g) => g.selectAll(".domain").remove()); // Add a g container for each group const g = svg .append("g") .attr("text-anchor", "end") .style("font", "14px sans-serif") .selectAll() .data(dataGrouped) .join("g") .attr("transform", ([key]) => `translate(0,${y(key)})`); // Append a grey line to each container. g.append("line") .attr("stroke", "#aaa") .attr("stroke-width", 3) .attr("x1", ([, values]) => x(d3.min(values, (d) => d.earnings))) .attr("x2", ([, values]) => x(d3.max(values, (d) => d.earnings))); // Append the dots to each container. g.append("g") .selectAll() .data(([, values]) => values) .join("circle") .attr("cx", (d) => x(d.earnings)) .attr("fill", (d) => color(d.percentile)) .attr("stroke", "white") .attr("stroke-width", "2") .attr("r", 9); // Append the label to each container. g.append("text") .attr("dy", "0.35em") .attr("x", () => x(xMin) - 6) .text(([group, values]) => group); // Expose the color scale for a legend. return Object.assign(svg.node(), { scales: { color } }); }34:Tde0,chartInstMargins = { // convert the data into a longform and filter only selected year postgrad const data = convertToLongform(selectInstitutionAndMarginsData).filter( (e) => e.yearPostgrad == yearInputInst ); // have this lookup so that the sort order can be defined const rowKeysLookup = new Map( data.map((d) => { const result = generateRowKeyLookupInst(d); return [result.id, result.label]; }) ); const sortedRowKeys = [...rowKeysLookup.keys()].sort(); const dataGrouped = d3.group(data, (d) => generateRowKeysInst(d)); const percentiles = new Set(data.map((d) => d.percentile)); // Specify the chart’s dimensions. const width = 928; const height = dataGrouped.size * 46; const marginTop = 30; const marginRight = 100; const marginBottom = 10; const marginLeft = 450; // scale values const min = d3.min(data.map((e) => e.earnings)); const max = d3.max(data.map((e) => e.earnings)); // Prepare the scales for positional and color encodings const x = d3 .scaleLinear() .domain([min, max]) .rangeRound([marginLeft, width - marginRight]) .nice(); const [xMin, xMax] = x.domain(); const y = d3 .scalePoint() .domain(sortedRowKeys.map((k) => rowKeysLookup.get(k))) .rangeRound([marginTop, height - marginBottom]) .padding(1); const color = d3 .scaleOrdinal() .domain(percentiles) .range(d3.schemeTableau10) .unknown("#ccc"); // Create the SVG container. const svg = d3 .create("svg") .attr("width", width) .attr("height", height) .attr("viewBox", [0, 0, width, height]) .attr("style", "max-width: 100%; height: auto;"); // Create the x axis. svg .append("g") .attr("transform", `translate(0,${marginTop})`) .call( d3 .axisTop(x) .ticks(7) .tickFormat((d) => "$" + d3.format(".2s")(d)) ) .call((g) => g .append("text") .text("earnings →") .attr("fill", "currentColor") .attr("transform", `translate(${width - marginRight},0)`) .attr("text-anchor", "end") .attr("dy", -22) ) .call((g) => g .selectAll(".tick line") .clone() .attr("stroke-opacity", 0.1) .attr("y2", height - marginBottom) ) .call((g) => g.selectAll(".domain").remove()); // Add a g container for each group const g = svg .append("g") .attr("text-anchor", "end") .style("font", "14px sans-serif") .selectAll() .data(dataGrouped) .join("g") .attr("transform", ([key]) => `translate(0,${y(key)})`); // Append a grey line to each container. g.append("line") .attr("stroke", "#aaa") .attr("stroke-width", 3) .attr("x1", ([, values]) => x(d3.min(values, (d) => d.earnings))) .attr("x2", ([, values]) => x(d3.max(values, (d) => d.earnings))); // Append the dots to each container. g.append("g") .selectAll() .data(([, values]) => values) .join("circle") .attr("cx", (d) => x(d.earnings)) .attr("fill", (d) => color(d.percentile)) .attr("fill-opacity", (d) => (d.earnings ? "1" : "0")) .attr("stroke", "white") .attr("stroke-width", "2") .attr("stroke-opacity", (d) => (d.earnings ? "1" : "0")) .attr("r", 9); // Append the label to each container. g.append("text") .attr("dy", "0.35em") .attr("x", () => x(xMin) - 6) .text(([group, values]) => group); // Expose the color scale for a legend. return Object.assign(svg.node(), { scales: { color } }); }35:Tdf5,chartInstSingle = { // convert the data into a longform and filter only selected year postgrad const data = convertToLongform(earningsTexas).filter( (e) => e.institution_label == selectInstitution ); // have this lookup so that the sort order can be defined const values = { y1: "1 year postgrad", y5: "5 years postgrad", y10: "10 years postgrad" }; const rowKeysLookup = new Map(Object.entries(values)); const sortedRowKeys = [...rowKeysLookup.keys()]; const dataGrouped = d3.group(data, (d) => d.yearPostgrad); const percentiles = new Set(data.map((d) => d.percentile)); // Specify the chart’s dimensions. const width = 928; const height = dataGrouped.size * 46; const marginTop = 30; const marginRight = 100; const marginBottom = 10; const marginLeft = 450; // scale values const min = d3.min(data.map((e) => e.earnings)); const max = d3.max(data.map((e) => e.earnings)); // Prepare the scales for positional and color encodings const x = d3 .scaleLinear() .domain([min, max]) .rangeRound([marginLeft, width - marginRight]) .nice(); const [xMin, xMax] = x.domain(); const y = d3 .scalePoint() .domain(sortedRowKeys.map((k) => rowKeysLookup.get(k))) .rangeRound([marginTop, height - marginBottom]) .padding(1); const color = d3 .scaleOrdinal() .domain(percentiles) .range(d3.schemeTableau10) .unknown("#ccc"); // Create the SVG container. const svg = d3 .create("svg") .attr("width", width) .attr("height", height) .attr("viewBox", [0, 0, width, height]) .attr("style", "max-width: 100%; height: auto;"); // Create the x axis. svg .append("g") .attr("transform", `translate(0,${marginTop})`) .call( d3 .axisTop(x) .ticks(7) .tickFormat((d) => "$" + d3.format(".2s")(d)) ) .call((g) => g .append("text") .text("earnings →") .attr("fill", "currentColor") .attr("transform", `translate(${width - marginRight},0)`) .attr("text-anchor", "end") .attr("dy", -22) ) .call((g) => g .selectAll(".tick line") .clone() .attr("stroke-opacity", 0.1) .attr("y2", height - marginBottom) ) .call((g) => g.selectAll(".domain").remove()); // Add a g container for each group const g = svg .append("g") .attr("text-anchor", "end") .style("font", "14px sans-serif") .selectAll() .data(dataGrouped) .join("g") .attr("transform", ([key]) => `translate(0,${y(rowKeysLookup.get(key))})`); // Append a grey line to each container. g.append("line") .attr("stroke", "#aaa") .attr("stroke-width", 3) .attr("x1", ([, values]) => x(d3.min(values, (d) => d.earnings))) .attr("x2", ([, values]) => x(d3.max(values, (d) => d.earnings))); // Append the dots to each container. g.append("g") .selectAll() .data(([, values]) => values) .join("circle") .attr("cx", (d) => x(d.earnings)) .attr("fill", (d) => color(d.percentile)) .attr("fill-opacity", (d) => (d.earnings ? "1" : "0")) .attr("stroke", "white") .attr("stroke-width", "2") .attr("stroke-opacity", (d) => (d.earnings ? "1" : "0")) .attr("r", 9); // Append the label to each container. g.append("text") .attr("dy", "0.35em") .attr("x", () => x(xMin) - 6) .text(([group, values]) => rowKeysLookup.get(group)); // Expose the color scale for a legend. return Object.assign(svg.node(), { scales: { color } }); }36:T4af,## Final Word Margins are wildly useful in the context of the PSEO data and applications. They allow us to: 1. describe an average/generic grad 2. walk the data 3. compare grad data against the average/generic grad 4. compare detailed programs against their parent (e.g. computer engineer vs engineering) 5. compare a generic state grad vs national (e.g. Texas vs US) 6. isolate large scale changes at a specific time (e.g. outcomes for grads during the great recession 2007-2009 vs everyone else) Some of these changes would slot easily into the existing implementation of [PSEO Explorer](https://lehd.ces.census.gov/applications/pseo/) while others would not. Publishing margins within the PSEO data product would likely require a different approach and application because using the data properly requires additional context. If we say, for example, that the average BA graduate "**earns $50k per year**" we'd certainly need to say "***based on data from X institution, Y states, and from 2002-2020**" (among many other things). We're exploring whether and how this can be accomplished. Stay tuned and email us anytime at [CES.PSEO.Feedback@census.gov](mailto:CES.PSEO.Feedback@census.gov) 37:T15f1,--- ## Creating the margins The following snippet shows how the y1 margins were created from the existing public use data files. y5, y10, and ipeds counts vars were excluded for brevity. They're essentially the exact same operation but on different variables. > Note: these margins ignore any status flags != 1 (aka are only calculated on "good" data) ```sql -- take the original pseoe table as used by explore and aggregate it to the institution level -- this means commenting out anything related to the institution -- agg_level_pseo, -- inst_level, -- institution, -- institution_label, -- institution_statefips, -- institution_statefips_label, -- create an empty table that's identical to the pseoe DROP TABLE IF EXISTS agg.earnings_all_inst; CREATE TABLE agg.earnings_all_inst (LIKE earnings.data INCLUDING ALL); WITH main as ( SELECT DISTINCT degree_level, cip_level, cipcode, grad_cohort, grad_cohort_years, geo_level, geography, ind_level, industry, degree_level_label, cipcode_label, grad_cohort_label, grad_cohort_id FROM earnings.data ), y1 AS ( SELECT -- agg_level_pseo, -- inst_level, -- institution, degree_level, cip_level, cipcode, grad_cohort, grad_cohort_years, geo_level, geography, ind_level, industry, (SUM(y1_p25_earnings * y1_grads_earn::bigint) / SUM(y1_grads_earn::bigint))::int AS y1_p25_earnings, (SUM(y1_p50_earnings * y1_grads_earn::bigint) / SUM(y1_grads_earn::bigint))::int AS y1_p50_earnings, (SUM(y1_p75_earnings * y1_grads_earn::bigint) / SUM(y1_grads_earn::bigint))::int AS y1_p75_earnings, (SUM(y1_grads_earn::bigint))::int AS y1_grads_earn, -- y5_p25_earnings, -- y5_p50_earnings, -- y5_p75_earnings, -- y5_grads_earn, -- y10_p25_earnings, -- y10_p50_earnings, -- y10_p75_earnings, -- y10_grads_earn, -- y1_ipeds_count, -- y5_ipeds_count, -- y10_ipeds_count, status_y1_earnings, status_y1_grads_earn, -- status_y5_earnings, -- status_y5_grads_earn, -- status_y10_earnings, -- status_y10_grads_earn, -- status_y1_ipeds_count, -- status_y5_ipeds_count, -- status_y10_ipeds_count, degree_level_label, cipcode_label, -- institution_label, grad_cohort_label, grad_cohort_id -- institution_statefips, -- institution_statefips_label FROM earnings.data WHERE status_y1_earnings = 1 AND status_y1_grads_earn = 1 GROUP BY -- agg_level_pseo, -- inst_level, -- institution, degree_level, cip_level, cipcode, grad_cohort, grad_cohort_years, geo_level, geography, ind_level, industry, -- y1_p25_earnings, -- y1_p50_earnings, -- y1_p75_earnings, -- y1_grads_earn, -- y5_p25_earnings, -- y5_p50_earnings, -- y5_p75_earnings, -- y5_grads_earn, -- y10_p25_earnings, -- y10_p50_earnings, -- y10_p75_earnings, -- y10_grads_earn, -- y1_ipeds_count, -- y5_ipeds_count, -- y10_ipeds_count, status_y1_earnings, status_y1_grads_earn, -- status_y5_earnings, -- status_y5_grads_earn, -- status_y10_earnings, -- status_y10_grads_earn, -- status_y1_ipeds_count, -- status_y5_ipeds_count, -- status_y10_ipeds_count, degree_level_label, cipcode_label, -- institution_label, grad_cohort_label, grad_cohort_id -- institution_statefips, -- institution_statefips_label ), --...etc for y5/y10/ipeds INSERT INTO agg.earnings_all_inst -- the actual query, filled with constants SELECT 'ZZ' as agg_level_pseo, '00' as inst_level, '00' as institution, main.degree_level, main.cip_level, main.cipcode, main.grad_cohort, main.grad_cohort_years, main.geo_level, main.geography, main.ind_level, main.industry, y1.y1_p25_earnings, y1.y1_p50_earnings, y1.y1_p75_earnings, y1.y1_grads_earn, y5.y5_p25_earnings, y5.y5_p50_earnings, y5.y5_p75_earnings, y5.y5_grads_earn, y10.y10_p25_earnings, y10.y10_p50_earnings, y10.y10_p75_earnings, y10.y10_grads_earn, y1_ipeds.y1_ipeds_count, y5_ipeds.y5_ipeds_count, y10_ipeds.y10_ipeds_count, y1.status_y1_earnings, y1.status_y1_grads_earn, y5.status_y5_earnings, y5.status_y5_grads_earn, y10.status_y10_earnings, y10.status_y10_grads_earn, y1_ipeds.status_y1_ipeds_count, y5_ipeds.status_y5_ipeds_count, y10_ipeds.status_y10_ipeds_count, main.degree_level_label, main.cipcode_label, 'All Institutions' as institution_label, main.grad_cohort_label, main.grad_cohort_id, '00' as institution_statefips, 'All States' as institution_statefips_label FROM main LEFT JOIN y1 USING(degree_level, cip_level, cipcode, grad_cohort, grad_cohort_years, geo_level, geography, ind_level, industry) -- ...etc for y10/ipeds vars ```38:[{"id":"05e39869b7803ba3e7a1b27976267ff7755ee91eca3d2e53a17cd739cd70a58db5b1bf6f88811b0a1089f93f0ba118e9e945336031db46a2ee8dd844e1ea9f97","href":"https://static.observableusercontent.com/files/05e39869b7803ba3e7a1b27976267ff7755ee91eca3d2e53a17cd739cd70a58db5b1bf6f88811b0a1089f93f0ba118e9e945336031db46a2ee8dd844e1ea9f97","name":"aud@3.csv","type":"text/csv","lastModified":1695314490257.161,"size":1169},{"id":"5e253684a2316c639dd4d5f2b895e0ced69dc9db4d8add7c88afaa3abb2cd9ddfcbfc3a03b8847e72e01a167b58c4b8f64138459c609b6a0b4f70e35a6c9ea17","href":"https://static.observableusercontent.com/files/5e253684a2316c639dd4d5f2b895e0ced69dc9db4d8add7c88afaa3abb2cd9ddfcbfc3a03b8847e72e01a167b58c4b8f64138459c609b6a0b4f70e35a6c9ea17","name":"aud_tx.csv","type":"text/csv","lastModified":1695314490257.161,"size":1397},{"id":"ae5d09c296e4333b118f143e6088762aedb33398f7a2d266983dc5e2f52f168ea519e237e96ad94c1cafa6bebadc5e2442afcf67e25336c3585b43b2551f6dec","href":"https://static.observableusercontent.com/files/ae5d09c296e4333b118f143e6088762aedb33398f7a2d266983dc5e2f52f168ea519e237e96ad94c1cafa6bebadc5e2442afcf67e25336c3585b43b2551f6dec","name":"aud_time.csv","type":"text/csv","lastModified":1695314490257.161,"size":741}] 2f:{"id":"645590ed20692432","slug":null,"title":"PSEO Margins Overview","thumbnail":"785f6f1f0c44b595639bd08c77287345446d224609abd8dd74db3d844e2077d1","access_level":"public","create_time":"$D2023-09-21T16:41:30.257Z","update_time":"$D2023-09-25T16:53:21.373Z","view_time":null,"first_public_time":"$D2023-09-21T16:41:30.257Z","has_unpublished_edits":null,"fork":{"id":"a33759822b5a6d59","slug":null,"title":"PSEO Margins User Story","owner":{"id":"3f0a5555946a91ca","login":"jodyhoonstarr","name":"Jody Hoon-Starr"}},"stars":0,"starred":false,"owner":{"type":"team","id":"3f0a5555946a91ca","login":"jodyhoonstarr","name":"Jody Hoon-Starr","avatar_url":"https://avatars-next.observableusercontent.com/f014ea613ff43e463687094066d9acde784357d91102e4a80afaee7d02ccf3fc","tier":"starter_2024","bio":""},"version":870,"publish_version":null,"role":null,"authored":false,"authors":[{"type":"individual","id":"741cd9e9458a548a","login":"jodyhoonstarr","name":"Jody Hoon-Starr","avatar_url":"https://avatars-next.observableusercontent.com/f014ea613ff43e463687094066d9acde784357d91102e4a80afaee7d02ccf3fc","tier":null,"bio":""}],"members":[{"type":"individual","id":"741cd9e9458a548a","login":"jodyhoonstarr","name":"Jody Hoon-Starr","avatar_url":"https://avatars-next.observableusercontent.com/f014ea613ff43e463687094066d9acde784357d91102e4a80afaee7d02ccf3fc","tier":null,"role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":0,"collection":null,"type":"notebook","body":{"titling":"auto","title":"PSEO Margins Overview","cells":[{"id":0,"value":"# PSEO Margins Overview","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":819,"value":"$31","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":638,"value":"
","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":587,"value":"## Disclaimer\n\nAll data shown here was generated with public data files. It is conceptually similar to what the production of PSEO margins would look like but should not be used in practice due to *composition effects* (*and any bugs that may have been introduced here*)","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":624,"value":"

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":631,"value":"## Margin Basics\n\nUse the existing published data and do a \"roll up\" \n\ne.g.\n- All Bachelor's engineering graduates \n- All Master's graduates in Texas","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":710,"value":"

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":690,"value":"## Basic Margin Example\n\nEarnings for all grads with: \nDegree Level: **${degreeLevel}** \nDetailed Program: **Communication Disorders Sciences and Services** (CIP 5102) \nState: **Any** \nInstitution: **Any** \nCohort: **Any** \nYears Postgrad: **${yearInput.replace(\"y\",\"\")}**\n\nBased on data from **${totalGradCountSimpleExample}** grads \n\n---\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":684,"value":"$32","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":723,"value":"Legend(chart.scales.color, { title: \"Earnings Percentile\", tickSize: 0 })","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":634,"value":"---\n

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":22,"value":"\n\n\n## Multiple Levels\n\nUsing the margins to \"walk\" the PSEO data. Each step below would have published data (earnings/flows). e.g.\n\n1. **Bachelor's Degree**\n2. Bachelor's Degree in **Engineering**\n3. Bachelor's Degree in **Computer Engineering**\n4. Bachelor's Degree in Computer Engineering for **All Institutions in Texas**\n5. Bachelor's Degree in Computer Engineering at the **University of Texas**(\\*)\n\n**OR**\n\n1. **All Institutions in Texas** (*special case - no margins on degree level*)\n2. **Bachelor's Degrees** for All Institutions in Texas\n3. Bachelor's Degrees in **Engineering** for All Institutions in Texas\n4. Bachelor's Degrees in **Computer Engineering** for All Institutions in Texas\n5. Bachelor's Degree in Computer Engineering at the **University of Texas**(\\*)\n\n(*) this is what we currently publish","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":673,"value":"

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":669,"value":"## Start exploring the data from:\n- Degree Level\n- State\n- Program (2-digit CIP)\n- Detailed Program (4-digit CIP)\n- Cohort (e.g. data for all BA graduates from 2007-2009)\n\nMargins let us mix and match (*instead of always going to a single institution*)","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":648,"value":"

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":646,"value":"## Added Context\n\nUse the margins to provide context while drilling into the data. e.g. data for\n\n- All **PhD** Grads\n- All PhD Grads in **Health Professions**\n- All PhD Grads in **Communication Disorders** (within Health Professions)\n- PhD Grads in Communication Disorders only in **Texas**","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":738,"value":"

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":339,"value":"## Data: Margins\nEarnings for Grads with a ${degreeLevel} degree\n\n---\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":85,"value":"viewof yearInput = Inputs.radio([\"y1\", \"y5\", \"y10\"], {\n value: \"y5\",\n label: \"Years Postgrad\"\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":90,"value":"$33","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":284,"value":"Legend(chart.scales.color, { title: \"Earnings Percentile\", tickSize: 0 })","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":742,"value":"---\n

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":351,"value":"---\n## Data: Institutions vs Margins\n\nEarnings for Grads with a ${degreeLevel} degree in Communication Disorders Sciences and Services\n\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":367,"value":"> *NOTE: the ability to toggle between institutions (as seen below) **WOULD NOT** be available in any Census application. It is for **testing and evaluation only**.*","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":358,"value":"viewof selectInstitution = Inputs.select([].concat(texasInstitutions), {\n label: \"Institution in Texas\"\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":403,"value":"viewof yearInputInst = Inputs.radio([\"y1\", \"y5\", \"y10\"], {\n value: \"y5\",\n label: \"Years Postgrad\"\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":390,"value":"$34","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":409,"value":"Legend(chartInstMargins.scales.color, {\n title: \"Earnings Percentile\",\n tickSize: 0\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":788,"value":"---\n

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":419,"value":"---\n## Data: Institution Only\n\n**${selectInstitution}**\n${degreeLevel} degrees\nin Communication Disorders Sciences and Services\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":422,"value":"$35","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":463,"value":"Legend(chartInstSingle.scales.color, {\n title: \"Earnings Percentile\",\n tickSize: 0\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":849,"value":"\n---\n

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":852,"value":"$36","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":865,"value":"---\n

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":863,"value":"## Changes Over Time\nAdditional experiments with regards to margins and changes over time can be seen in the notebook: [PSEO Margins Over Time](https://observablehq.com/d/14fc977cede6296c)","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":804,"value":"---\n

\n

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":42,"value":"# Appendix\n---\n## Data used\n\nThis data has been generated using the existing public use PSEO data files. The `pseoe_all` file was aggregated to \n- `Degree Level x All Institutions`\n- `Degree Level x All Institutions x State`\n\nTwo attached files contain the information used here\n- `aud.csv` - margins for CIP 5102, its parent CIP 51, and CIP 00 All\n- `aud_tx.csv` - standard pseoe data for Texas CIP 5102\n- `aud_time.csv` - are margins over time, aka grad_cohort ! '0000'\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":480,"value":"$37","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":33,"value":"---\n## Helper functions","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":9,"value":"earningsMargins = d3.csvParse(await FileAttachment(\"aud@3.csv\").text(), (d) => {\n return {\n ...d,\n y1_p25_earnings: d.y1_p25_earnings ? +d.y1_p25_earnings : null,\n y1_p50_earnings: d.y1_p50_earnings ? +d.y1_p50_earnings : null,\n y1_p75_earnings: d.y1_p75_earnings ? +d.y1_p75_earnings : null,\n y1_grads_earn: d.y1_grads_earn ? +d.y1_grads_earn : null,\n y5_p25_earnings: d.y5_p25_earnings ? +d.y5_p25_earnings : null,\n y5_p50_earnings: d.y5_p50_earnings ? +d.y5_p50_earnings : null,\n y5_p75_earnings: d.y5_p75_earnings ? +d.y5_p75_earnings : null,\n y5_grads_earn: d.y5_grads_earn ? +d.y5_grads_earn : null,\n y10_p25_earnings: d.y10_p25_earnings ? +d.y10_p25_earnings : null,\n y10_p50_earnings: d.y10_p50_earnings ? +d.y10_p50_earnings : null,\n y10_p75_earnings: d.y10_p75_earnings ? +d.y10_p75_earnings : null,\n y10_grads_earn: d.y10_grads_earn ? +d.y10_grads_earn : null\n };\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":70,"value":"earningsTexas = d3.csvParse(await FileAttachment(\"aud_tx.csv\").text(), (d) => {\n return {\n ...d,\n y1_p25_earnings: d.y1_p25_earnings ? +d.y1_p25_earnings : null,\n y1_p50_earnings: d.y1_p50_earnings ? +d.y1_p50_earnings : null,\n y1_p75_earnings: d.y1_p75_earnings ? +d.y1_p75_earnings : null,\n y1_grads_earn: d.y1_grads_earn ? +d.y1_grads_earn : null,\n y5_p25_earnings: d.y5_p25_earnings ? +d.y5_p25_earnings : null,\n y5_p50_earnings: d.y5_p50_earnings ? +d.y5_p50_earnings : null,\n y5_p75_earnings: d.y5_p75_earnings ? +d.y5_p75_earnings : null,\n y5_grads_earn: d.y5_grads_earn ? +d.y5_grads_earn : null,\n y10_p25_earnings: d.y10_p25_earnings ? +d.y10_p25_earnings : null,\n y10_p50_earnings: d.y10_p50_earnings ? +d.y10_p50_earnings : null,\n y10_p75_earnings: d.y10_p75_earnings ? +d.y10_p75_earnings : null,\n y10_grads_earn: d.y10_grads_earn ? +d.y10_grads_earn : null\n };\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":515,"value":"earningsMarginsOverTime = d3.csvParse(\n await FileAttachment(\"aud_time.csv\").text(),\n (d) => {\n return {\n ...d,\n y1_p25_earnings: d.y1_p25_earnings ? +d.y1_p25_earnings : null,\n y1_p50_earnings: d.y1_p50_earnings ? +d.y1_p50_earnings : null,\n y1_p75_earnings: d.y1_p75_earnings ? +d.y1_p75_earnings : null,\n y1_grads_earn: d.y1_grads_earn ? +d.y1_grads_earn : null,\n y5_p25_earnings: d.y5_p25_earnings ? +d.y5_p25_earnings : null,\n y5_p50_earnings: d.y5_p50_earnings ? +d.y5_p50_earnings : null,\n y5_p75_earnings: d.y5_p75_earnings ? +d.y5_p75_earnings : null,\n y5_grads_earn: d.y5_grads_earn ? +d.y5_grads_earn : null,\n y10_p25_earnings: d.y10_p25_earnings ? +d.y10_p25_earnings : null,\n y10_p50_earnings: d.y10_p50_earnings ? +d.y10_p50_earnings : null,\n y10_p75_earnings: d.y10_p75_earnings ? +d.y10_p75_earnings : null,\n y10_grads_earn: d.y10_grads_earn ? +d.y10_grads_earn : null\n };\n }\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":705,"value":"totalGradCountSimpleExample = convertToLongformTimeSeries(\n earningsMargins\n).filter(\n (e) =>\n e.yearPostgrad === yearInput &&\n e.cipcode === \"5102\" &&\n e.institution_statefips === \"00\"\n)[0].grads","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":557,"value":"earningsMarginsOverTimeLongform = convertToLongformTimeSeries(\n earningsMarginsOverTime\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":354,"value":"texasInstitutions = [...new Set(earningsTexas.map((e) => e.institution_label))]","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":383,"value":"selectInstitutionAndMarginsData = [\n ...earningsMargins.filter((d) => d.cipcode == \"5102\"),\n ...earningsTexas.filter((d) => d.institution_label == selectInstitution)\n]","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":199,"value":"convertToLongform = (data) =>\n data.flatMap((d) => {\n return [\"y1\", \"y5\", \"y10\"].flatMap((y) => {\n return [\"p25\", \"p50\", \"p75\"].flatMap((p) => {\n return {\n institution: d.institution,\n institution_label: d.institution_label,\n institution_statefips: d.institution_statefips,\n institution_statefips_label: d.institution_statefips_label,\n cipcode: d.cipcode,\n cipcode_label: d.cipcode_label,\n degree_level: d.degree_level,\n degree_level_label: d.degree_level_label,\n grad_cohort: d.grad_cohort,\n grad_cohort_label: d.grad_cohort_label,\n yearPostgrad: y,\n percentile: p,\n earnings: d[`${y}_${p}_earnings`]\n };\n });\n });\n })","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":520,"value":"convertToLongformTimeSeries = (data) =>\n data.flatMap((d) => {\n return [\"y1\", \"y5\", \"y10\"].flatMap((y) => {\n return {\n institution: d.institution,\n institution_label: d.institution_label,\n institution_statefips: d.institution_statefips,\n institution_statefips_label: d.institution_statefips_label,\n cipcode: d.cipcode,\n cipcode_label: d.cipcode_label,\n degree_level: d.degree_level,\n degree_level_label: d.degree_level_label,\n grad_cohort: d.grad_cohort,\n grad_cohort_label: d.grad_cohort_label,\n yearPostgrad: y,\n grads: d[`${y}_grads_earn`]\n };\n });\n })","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":24,"value":"function getDistinctValues(arr, property) {\n const distinctValues = new Set();\n\n arr.forEach((obj) => {\n if (obj.hasOwnProperty(property)) {\n distinctValues.add(obj[property]);\n }\n });\n\n return Array.from(distinctValues);\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":172,"value":"generateRowKeys = (d) => `${d.cipcode_label} (${d.institution_statefips_label})`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":297,"value":"generateRowKeyLookup = (d) => ({\n id: `${d.institution_statefips}-${d.cipcode}`,\n label: generateRowKeys(d)\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":395,"value":"generateRowKeysInst = (d) =>\n `${d.institution_label} (${d.institution_statefips_label})`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":392,"value":"generateRowKeyLookupInst = (d) => ({\n id: `${d.institution_statefips}-${d.institution}`,\n label: generateRowKeysInst(d)\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":27,"value":"degreeLevel = getDistinctValues(earningsMargins, \"degree_level_label\")[0]","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":31,"value":"gradCohort = getDistinctValues(earningsMargins, \"grad_cohort_label\")[0]","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":286,"value":"import { Legend } from \"@d3/color-legend\"","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W38","stdlib":"1"}} 30:"$2f:body"