Inference in differences-in-differences with few treated groups and heteroskedasticity

Differences-in-Differences (DID) is one of the most widely used identification strategies in applied economics. However, how to draw inferences in DID models when there are few treated groups remains an open question. We show that the usual inference methods used in DID models might not perform well when there are few treated groups and errors are heteroskedastic. In particular, we show that when there is variation in the number of observations per group, inference methods designed to work when there are few treated groups tend to (under-) over-reject the null hypothesis when the treated groups are (large) small relative to the control groups. This happens because larger groups tend to have lower variance, generating heteroskedasticity in the group x time aggregate DID model. We provide evidence from Monte Carlo simulations and from placebo DID regressions with the American Community Survey (ACS) and the Current Population Survey (CPS) datasets to show that this problem is relevant even in datasets with large numbers of observations per group. We then derive an alternative inference method that provides accurate hypothesis testing in situations where there are few treated groups (or even just one) and many control groups in the presence of heteroskedasticity. Our method assumes that we can model the heteroskedasticity of a linear combination of the errors. We show that this assumption can be satisfied without imposing strong assumptions on the errors in common DID applications. With many pre-treatment periods, we show that this assumption can be relaxed. Instead, we provide an alternative inference method that relies on strict stationarity and ergodicity of the time series. Finally, we consider two recent alternatives to DID when there are many pre-treatment periods. We extend our inference methods to linear factor models when there are few treated groups. We also derive conditions under which a permutation test for the synthetic control estimator proposed by Abadie et al. (2010) is robust to heteroskedasticity and propose a modification on the test statistic that provided a better heteroskedasticity correction in our simulations.

Inference in differences-in-differences with few treated groups and heteroskedasticity

Differences-in-Differences (DID) is one of the most widely used identification strategies in applied economics. However, how to draw inferences in DID models when there are few treated groups remains an open question. We show that the usual inference methods used in DID models might not perform well when there are few treated groups and errors are heteroskedastic. In particular, we show that when there is variation in the number of observations per group, inference methods designed to work when there are few treated groups tend to (under-) over-reject the null hypothesis when the treated groups are (large) small relative to the control groups. This happens because larger groups tend to have lower variance, generating heteroskedasticity in the group x time aggregate DID model. We provide evidence from Monte Carlo simulations and from placebo DID regressions with the American Community Survey (ACS) and the Current Population Survey (CPS) datasets to show that this problem is relevant even in datasets with large numbers of observations per group. We then derive an alternative inference method that provides accurate hypothesis testing in situations where there are few treated groups (or even just one) and many control groups in the presence of heteroskedasticity. Our method assumes that we know how the heteroskedasticity is generated, which is the case when it is generated by variation in the number of observations per group. With many pre-treatment periods, we show that this assumption can be relaxed. Instead, we provide an alternative application of our method that relies on assumptions about stationarity and convergence of the moments of the time series. Finally, we consider two recent alternatives to DID when there are many pre-treatment groups. We extend our inference method to linear factor models when there are few treated groups. We also propose a permutation test for the synthetic control estimator that provided a better heteroskedasticity correction in our simulations than the test suggested by Abadie et al. (2010).

Biodistribution of samarium-153-EDTMP in rats treated with docetaxel

Purpose: Many patients with metastatic bone disease have to use radiopharmaceuticals associated with chemotherapy to relieve bone pain. The aim of this study was to assess the influence of docetaxel on the biodistribution of samarium-153-EDTMP in bones and other organs of rats. Methods: Wistar male rats were randomly allocated into 2 groups of 6 rats each. The DS (docetaxel/samarium) group received docetaxel (15 mg/kg) intraperitoneally in two cycles 11 days apart. The S (samarium/control) group rats were not treated with docetaxel. Nine days after chemotherapy, all the rats were injected with 0.1ml of samarium-153-EDTMP via orbital plexus (25μCi). After 2 hours, the animals were killed and samples of the brain, thyroid, lung, heart, stomach, colon, liver, kidney and both femurs were removed. The percentage radioactivity of each sample (% ATI/g) was determined in an automatic gamma-counter (Wizard-1470, Perkin-Elmer, Finland). Results: On the 9th day after the administration of the 2nd chemotherapy cycle, the rats had a significant weight loss (314.50±22.09g) compared (p<0.5) to pre-treatment weight (353.66± 22.8). The % ATI/g in the samples of rats treated with samarium-153-EDTMP had a significant reduction in the right femur, left femur, kidney, liver and lungs of animals treated with docetaxel, compared to the control rats. Conclusion: The combination of docetaxel and samarium-153-EDTMP was associated with a lower response rate in the biodistribution of the radiopharmaceutical to targeted tissues. Further investigation into the impact of docetaxel on biodistribution of samarium-153-EDTMP would complement the findings of this study

Biosorption of Cr(III) using in natura and chemically treated tropical peats

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Expression of matrix metalloproteinases, type IV collagen, and interleukin-10 in rabbits treated with morphine after lamellar keratectomy

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Transition metal rates in latosol twice treated with sewage sludge

A reciclagem agrícola do lodo de esgoto tem provocado o acúmulo de metais pesados no solo e na água, podendo atingir níveis tóxicos e causar danos às plantas cultivadas, aos animais e ao homem, por meio da cadeia trófica. Neste intuito foi desenvolvido o presente experimento, em condições de campo, entre 2000 e 2002, onde foram avaliados os efeitos da aplicação de lodo de esgoto por dois anos, sobre a extração de metais de transição (essenciais e não) pelo extrator DTPA em um Latossolo Vermelho distrófico (LVd) de textura média. As concentrações dos elementos metálicos: Mn, Fe, Cd, Ni, Co, Pb e Cr não foram detectados pelo método da absorção atômica na solução obtida com o extrator DTPA. A aplicação de lodo de esgoto causou inicialmente pequena elevação no pH do solo, posteriormente a diminuição do mesmo, e manteve-se próximo ao original. Foi possível concluir que, com a aplicação consecutiva do lodo, os teores extraíveis de Fe e Mn nas amostras de solos aumentaram gradativamente nos dois anos agrícolas, com as doses do lodo de esgoto aplicado, época de amostragens, e foram superiores ao tratamento testemunha. O extrator apresentou capacidade restrita para avaliação da fitodisponibilidade dos metais pesados decorrentes das baixas concentrações nas amostras de solo.

Histopathological biomarkers in pacu (Piaractus mesopotamicus) infected with aeromonas hydrophila and treated with antibiotics

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Activation and early parthenogenesis of bovine oocytes treated with ethanol and strontium

Efficient artificial activation is indispensable for the success of cloning programs. Strontium has been shown to effectively activate mouse oocytes for nuclear transfer procedures, however, there is limited information on its use for bovine oocytes. The present study had as objectives: (1) to assess the ability of strontium to induce activation and parthenogenetic development in bovine oocytes of different maturational ages in comparison with ethanol; and (2) to verify whether the combination of both treatments improves activation and parthenogenetic development rates. Bovine oocytes were in vitro matured for 24, 26, 28, and 30 h, and treated with ethanol (E, 7% for 5 min) or strontium chloride (S, 10 mM SrCl2 for 5 h) alone or in combination: ethanol + strontium (ES) and strontium + ethanol (SE). Activated oocytes were cultured in vitro in synthetic oviductal fluid (SOF) medium and assessed for pronuclear formation (15-16 h), cleavage (46-48 h) and development to the blastocyst stage (M). Treatment with ethanol and strontium promoted similar results regarding pronuclear formation (E, 20-66.7%; S, 26.7-53.3%; P > 0.05) and cleavage (E, 12.8-40.6%; S, 16.1-41.9%; P > 0.05), regardless of oocyte age. The actions of both strontium and ethanol were influenced by oocyte age: ethanol induced greater activation rates after 28 and 30 h of maturation (48.4 and 66.7% versus 20.0 and 23.3% for 24 and 26 It, respectively; P < 0.05) and strontium after 30 It (53.3%) was superior to 24 and 26h (26.7% for both). Blastocyst development rates were minimal in all treatments (0.0-6.3%; P > 0.05), however, when the mean (+/-S.D.) cell number in blastocysts at the same maturational period was compared, strontium treatment was superior to ethanol for activation rates (82 +/- 5.7 and 89.5 +/- 7.8 versus 54 and 61, at 28 and 30 h, respectively). Improved results were obtained by combined treatments. The combination of ethanol and strontium resulted in similar pronuclear formation (ES, 36.7-83.9%; SE, 53.1-90.3%) and cleavage rates (ES, 31.3-81.3%; SE, 65.6-80.7%). Regarding embryo development, there was no difference (P > 0.05) between treatments, and blastocysts were only obtained in treatment SE at 24 and 26 h (6.5% for both). It is concluded that, SrCl2 induces activation and parthenogenetic development in bovine oocytes. (C) 2003 Elsevier B.V. All rights reserved.

Effect of timing of estradiol benzoate administration upon synchronization of ovulation in suckling Nelore cows (Bos indicus) treated with a progesterone-releasing intravaginal device

The present study investigated how the timing of the administration of estradiol benzoate (EB) impacted the synchronization of ovulation in fixed-time artificial insemination protocols of cattle. To accomplish this, two experiments were conducted, with EB injection occurring at different times: at withdrawal of the progesterone-releasing (N) intravaginal device or 24 h later. The effectiveness of these times was compared by examining ovarian follicular dynamics (Experiment 1, n = 30) and conception rates (Experiment 2, n = 504). In Experiment 1, follicular dynamics was performed in 30 Nelore cows (Bos indicus) allocated into two groups. on a random day of the estrous cycle (Day 0), both groups received 2 mg of EB i.m. and a P4-releasing intravaginal device, which was removed on Day 8, when 400 IU of eCG and 150 mu g of PGF were administered. The control group (G-EB9; n = 15) received 1 mg of EB on Day 9, while Group EB8 (G-EB8; n = 15) received the same dose a day earlier. Ovarian ultrasonographic evaluations were performed every 8 h after device removal until ovulation. The timing of EB administration (Day 8 compared with Day 9) did affect the interval between P4 device removal to ovulation (59.4 +/- 2.0 h compared with 69.3 +/- 1.7 h) and maximum diameter of dominant (1.54 +/- 0.06 a cm compared with 1.71 +/- 0.05 b cm, P = 0.03) and ovulatory (1.46 +/- 0.05 a cm compared with 1.58 +/- 0.04 b cm, P < 0.01) follicles. In Experiment 2,504 suckling cows received the same treatment described in Experiment 1, but insemination was performed as follows: Group EB8-AI48h (G-EB8-AI48h; n = 119) and Group EB8-AI54h (G-EB8-AI54h; n = 134) received 1 mg of EB on Day 8 and FrAI was performed, respectively, 48 or 54 h after P4 device removal. Group EB9-AI48h (G-EB9-AI48h; n = 126) and Group EB9-AI54h (G-EB9-AI54h n = 125) received the same treatments and underwent the same FTAI protocols as G-EB8-AI48h and G-EB8-AI54h, respectively; however, EB was administered on Day 9. Conception rates were greater (P < 0.05) in G-EB9-AI54h 163.2% (79/125) a], G-EB9-AI48h [58.7% (74/126) a] and G-EB8-AI48h [58.8% (70/119) a] than in G-EB8-AI54h [34.3% (46/134) b]. We concluded that when EB administration occurred at device withdrawal (D8), the interval to ovulation shortened and dominant and ovulatory follicle diameters decreased. Furthermore, when EB treatment was performed 24 h after device removal, FTAI conducted at either 48 or 54 h resulted in similar conception rates. However, EB treatment on the same day as device withdrawal resulted in a lesser conception rate when FTAI was conducted 54 h after device removal. (C) 2007 Elsevier B.V. All rights reserved.

Chemical properties and enzyme activity in a sewage sludge-treated soil

A soil sample was taken from the top 0-20cm at Jaboticabal county, São Paulo State, Brazil, air dried, sieved to 5mm, and placed into pots (2700g per pot). Sewage sludge was air-dried, ground to 2mm, and thoroughly mixed to the top 0-10cm soil of each pot, which were irrigated with distilled water in a total volume equivalent to the last 30years average rainfall in the region. Sorghum was sowed 120days after sewage sludge incorporation and then the irrigation was made according to the plants' requirement. When the plants were about 10 cm high, they were thinned to two per pot. Soil samples (0-10, 10-20, and 20-30 cm depth) were obtained immediately after the incorporation of sewage sludge and at 30, 60, 120, and 170 days after, air dried, sieved to 2 mm and analyzed for organic matter (OM), pH (0,01 mol L-1 CaCl2), extractable P (resin), potassium (K), calcium (Ca), and magnesium (Mg), amylase and cellulase activity. Sewage sludge increased soil OM, pH, extractable phosphorus (P), K. Ca. amylase and cellulase activity, especially at the rate 16 t ha(-1). Organic matter, extractable P, K, Ca, Mg. and amylase activity were higher in the top 0-10cm, while pH was higher in the 20-30cm layer. Amylase activity was not affected by sampling depth. Organic matter, pH, extractable P. K, Ca, and Mg decreased during the experimental period. Amylase activity decreased until sorghum was sowed and increased afterwards. Cellulase activity increased until 90 days after sewage sludge application and then decreased. Sewage sludge used in the experiment should already contain some amylase activity or a substance that was a soil enzyme activator and also a substance that was an inhibitor of soil cellulase inhibitor. Sonic of the plant nutrients contained in sewage sludge, mainly P, did not migrate down the soil column. an indication that sewage sludge should be incorporated into the soil to improve nutrient bioavailability. Sorghum roots increased amylase activity but did not affect cellulase activity.

Lead distribution in organic matter of loamy oxisol treated with manure and its effect on microbial biomass

This paper reports a study regarding the distribution of lead in the organic matter fractions of a loamy oxisol treated with stable manure, and its effect on the soil microbial biomass. The experimental design was a completely randomized factorial with three replicates. Treatments were four lead (Pb) rates (0, 200, 400, and 600 mg kg(-1) soil), two levels of manure (0 and 30 Mg ha(-1)) and two sampling times (30 and 60 days after lead application). Total and soil lead soluble in Mehlich 1 extractant, and Ph in the different fractions of the organic matter (fulvic acid, humic acid, and residual) were evaluated along with the soil microbial biomass. Data showed that lead, applied as PbCl2, was concentrated in the residual fraction. The application of manure at the rate of 30 Mg ha(-1) did not affect lead distribution in the fractions until 60 days after incorporation. Lead extracted by Mehlich 1 increased as function of the quantity applied. Manure stimulated the growth of the soil microbial biomass which was reduced by the 200 mg kg(-1) doses of lead, with manure application, increasing, however, with larger rates. Mehlich 1 extracted only part of the lead present in the residual fraction.

Longevity of Oncidium varicosum (Orchidaceae) inflorescences treated with 1-methylciclopropene

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Nickel in soil and maize plants grown on an oxisol treated over a long time with sewage sludge

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Ionized-air-treated curaua fibers as reinforcement for phenolic matrices

Curaua fibers were treated with ionized air to improve the fiber/phenolic matrix adhesion.The treatment with ionized air did not change the thermal stability of the fibers. The impact strength increased with increase in the fiber treatment time. SEM micrographs of the fibers showed that the ionized air treatment led to separation of the fiber bundles. Treatment for 12 h also caused a partial degradation of the fibers, which prompted the matrix to transfer the load to a poorer reinforcing agent during impact, thereby decreasing the impact strength of the related composite. The composites reinforced with fibers treated with ionized air absorbed less water than those reinforced with untreated fibers.

Use of Cassava Wastewater Treated Anaerobically with Alkaline Agents as Fertilizer for Maize (Zea mays L.)

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)